I'm a British Physician-Scientist in-training, building the healthcare technologies of tomorrow. 2021-05-13T00:00:00-00:00 https://souradip.com/ Souradip Mookerjee [email protected] 2021-05-13T00:00:00-00:00 https://souradip.com/blog/free-sequencing I should probably say that my primary reason for signing up to a stem cell donor registry is to actually be a donor! This data is generated as a byproduct of signing up and is also quite interesting on a personal level. I was thinking that framing it in a way to also find out interesting things about yourself might be a good way to encourage people who might not have thought about it to sign up for the first time, and this would be far more cost-effective than a traditional in-person donor recruitment drive. The dream of personalised medicine For the longest time, we have treated all people as the same in medicine. Everyone with a disease has the same disease, after all. But in reality, how that disease or condition affects someone's life depends on many factors, including their genetics. Some people's genes make them more likely to get certain diseases than others. If you know you specifically were more at risk of developing a certain disease, you could take preventative action to avoid it! For example, if you knew you were more susceptible to skin cancer, maybe you might take slightly more care in applying suncream than someone who is not. Genes encode for proteins. Drugs target proteins. So it makes sense that if your genes code for slightly different proteins to the ones the drug was designed for, it might not even work! A simple example of this is codeine. Some people swear by it as a painkiller but it makes no difference to others. This is because codeine is a prodrug that is broken down by the liver to its active form. Some people have a gene that means they can do this really well and feel the benefits, while some people have a different version of this gene that means the codeine never breaks down at all, and they stay unaffected. This is the dream of personalised medicine. To be able to take the unique features of our genomes into account when discovering how certain diseases or other environmental factors affect us, and how our treatments might be affected by our genes. Knowledge is power. Unfortunately, until now it's been relatively expensive to get your genome sequenced fully. Thankfully, you don't need to sequence everything - that includes all the genes that are fairly boring. They say 50% of your genes are the same as in a banana after all! What we want are the interesting genes. Even so, companies like 23andme charge quite a premium for their service, and may be doing anything with your data afterwards. Will it end up in the hands of insurance companies? But in this article, I will share with you how I got my own genome sequenced for free, learnt a bit more about myself while I was at it, made myself known in case I could save someone's life one day, and you can too! The role of HLA (Human Leukocyte Antigen) HLA (Human Leukocyte Antigens) are parts of our genome that are home to an incredible amoutn of variation from person to person. The purpose of HLA in our cells is to take whatever is inside of our cells (HLA Class I) or what they happen to take up from their enviroment (HLA Class II), chop them up, and present snippets of it to the immune system. We have several different HLA's in our genome. HLA-A, HLA-B, HLA-C and HLA-E are Class I (they show what's inside our cells). All of the HLA-D's are class II (they show what the cells have taken up from their environment). What snippets get shown to the immune system depends on your HLA type. If you have a HLA-A01, (type 01 of the HLA-A protein), you will show different snippets to the immune system to someone with HLA-A02. We have so much variation here because things that might infect us like viruses will also get chopped up and display themselves on here. This is how our immune system spots infected cells! And that's just on one HLA protein (also known as a locus, or specific point in our genome). Then you have to multiply that complexity with the number of different HLA molecules we have (A, B, C and so on). One individual will likely have hardly any matches to another person chosen at random. The variation present here means that viruses can't evolve to evade it. If it evolves to avoid one type, another type will be able to catch it. That also means certain HLA types are more susceptible to certain specific viruses more than other people. Because it also presents snippets of other proteins in our cells, this opens to door for autoimmunity. Some snippets of inside our cells trigger specific autoimmune diseases more than others. This is why some HLA types are specifically more prone to specific autoimmune diseases (e.g. HLA-B27 with ankylosing spondylosis and HLA-B51 with Behçets disease). Finally, because this is specifically what the immune system uses to check the identity of tissues, this is also what gets "matched" when someone has a transplant, to reduce the chance of rejection. Because it's unlikely that you could ever get someone with an absolutely perfect match (unless you have a twin!), people often still need some immunosuppressants after a transplanted organ. How I got my HLA typing info (and more) for free thanks to GDPR You may be aware of bone marrow, or stem cell transplants. These are life-saving for the people who need them, especially after a blood cancer. You can sign up for free at charities such as Anthony Nolan or DKMS in the UK. I would strongly recommend you to join, it's fast and easy (you only need to spit into a bottle!). You might be the only person in the world that could save someone's life. It doesn't matter which you join, as they are all linked together in a global database. They take your saliva, and test your DNA for the types of HLA you have in there, in order to be able to best match you with a recipient if it was ever needed. Under the recent EU GDPR Data Protection laws, you have the right to request what data any organisation stores about you. A massive win for privacy, a massive headache for many websites with little cookie banners all over. From my background in technology to notice that this doesn't just apply to websites, this applies to any organisation including the above stem cell registries. So I sent them a (polite) request for my data under this law and they replied! After sending in proof of my identity, they sent me back my genotyping results! This means that I basically got part of my genome sequenced for free (really, the most interesting part! I don't need to know about the bits I have in common with a banana after all!). And it was a win-win scenario - they get to call upon me if someone needs my stem cells (a painless procedure that's no more complicated than donating blood)! Making sense of it all What they sent me back was more than just my HLA type, it was also my KIR type (what is present on my NK cells, or natural killer cells - I wrote a whole essay on this if you're interested here!), my CCR5 status (the mutation which if present, confers immunity to HIV infections) and plenty more. My blood type (of course!) My status for IgG CMV This tells me if I've ever been infected by CMV in the distant past, a virus that is normally harmless and persists inside you not doing anything until you're immunosuppressed. Spread by saliva ahem. CCR5 status (sadly, I have no mutations here which would give me immunity against HIV...) KIR type HLA type The first three on this list are fairly easy to make sense of. The last two are the most interesting but also harder to make sense of. I would treat this as a bit of fun, not medical advice! Interpreting HLA types Certain HLA types can predispose you towards certain autoimmune conditions, while others can protect you. Often the studies in this area are correlational, but some work has been done expanding on specific HLA types in mouse models. Firstly, how do we read the HLA code they give us? It's all in the WHO Standard Nomenclature. To simplify things, you only need to know the first number - the latter ones are far too detailed to make sense of at the moment. When you see HLA-A*01:01:03G just read it as HLA-A*01. NB: Some journal websites find it difficult to display the asterisk and so they replace it with a w - so HLA-Cw04 and HLA-C*04 both mean the same thing. What I discovered As a quick example, I discovered the reason my granddad and I both have a bit of psoriasis is because I have the HLA-C*12 allele, which predisposes to this skin condition! The more I read up on my HLA types the more I learn. Today I discovered that my B35-C04 combo makes me at a high risk of rapid progression to AIDS if I ever catch HIV. One to avoid! Finally, I should also note that Coeliac disease is a T-cell mediated disease predominantly of people carrying HLA-DQ2 or HLA-DQ8. I have neither of these, so it seems that I am at very low risk of developing coeliac, and I can eat as much gluten as I'd like! Interpreting KIR types The KIR genes are involved in a very different part of the immune system, the natural killer cells rather than the lymphocytes. I won't go into too much here, but broadly the research splits people up into either Haplotype A or Haplotype B depending on which KIR genes you have. Haplotype A is associated with resistance to infections but a higher chance of pre-eclampsia, while Haplotype B is associated with increased reproductive fitness but more susceptibility to infections. Read more about this in my essay on the topic. :) The future Although this was a bit of fun, the ideas in this post open up an exciting world of consumer-level genetic profiling, where there are no data security concerns (ahem, 23andme), you get to help save other people's lives, and find out a bit about yourself in the process. There is even evidence that you might be able to predict your individual response to certain medicines. How cool is that? 2021-01-30T00:00:00-00:00 https://souradip.com/blog/experience-bame-person-working-in-hospital-covid I'm a doctor! (ish) Since last June, a lot of things happened. I finished my PhD and passed my viva! I'm now a doctor! That is, of the PhD kind. I'm still working on becoming a doctor of the medical kind, and I have one-and-a-half years left of Clinical School at Cambridge. I had a grant total of one day off between submitting my thesis and returning to Year 5(/6) of my medical course! Souradip Mookerjee @souramooOfficially a Dr? I just passed my (3hr) Zoom viva, I now have a PhD from Cambridge! #UniversityOfZoom #PhDone #MedTwitter #Science371 12:10 PM · Dec 18, 2020 Vaccinated Lent term came around, with only clinical students (including vets) were allowed to return to the University, so College has felt quite empty. I did manage to get myself vaccinated (first dose), and the hospital did a great job at rolling this out and not leaving students out of the queue. Souradip Mookerjee @souramooDid you even get the COVID vaccine if you didn't take a photo?? #CovidVaccine #PfizerBioNTech #COVID19 #MedStudentTwitter25 01:03 PM · Jan 14, 2021 The good parts This year, the specialist medicine placements I've been on have been quite possibly the most interesting of all of medical school so far. I've had the knowledge to really make sense of all the physiology I'm seeing in action, reading the X rays, blood test results and ECGs myself and solving these puzzles. It feels like all the previous years of medical school are finally paying off! Things that seemed impossible to understand and unconnected suddenly now snap into place, and I can see the interactions and logical connections between the mechanisms of disease and the resulting clinical picture. However, there was one thing that I was slightly dreading this term. Souradip Mookerjee @souramooclinical medicine starter pack 2021: *gulp*#COVID19 #Lockdown #MedTwitter #medstudenttwitter15 11:28 PM · Jan 4, 2021 That's right! I started on infectious diseases, and since then it's been a whistle-stop tour through everything from oncology, through to cardiology and respiratory medicine. I've been at the Royal Papworth hospital this last week and the teaching was incredibly well organised. I had the chance to see some truly cutting-edge treatments from pulmonary endarterectomies through to ECMO! I even got the chance to intubate someone under supervision before an open heart operation. I felt like a real medical student, turning up to the wards with a friend, looking over patient notes and results, figuring out a management plan and getting taught from the junior doctors on the ward. I've also been added to the staff bank at the hospital to cover shifts because they're really struggling with having enough people to run a safe service. My name COVID has meant we have to wear quite strict PPE everywhere in the hospital. But in a way it feels somewhat dehumanising. It's really hard to talk with an FFP3 mask on. This has become a problem because it's hard to introduce myself as I normally would. Not only is my face covered up (and so it's easy to just become anonymous, with nobody recognising me between days, making building long-term rapport with other staff hard), but this is a conversation I've had a lot with other healthcare staff: "Hello, I'm Souradip and I'm a medical student here" "Hi, sorry, what did you say your name was?" "Souradip" "So- who?" "Souradip, S-O-U-R-A-D-I-P" "..." "It's like sour dip with an a in the middle" (my go-to joke to lighten the mood) "What?" Not only is it hard for people to know my face, they can't even know my name! And humour doesn't seem to translate well when they can't hear half of the joke. And the best way to kill a joke is to keep repeating it again and again... I quickly threw together an app so that I can pull out my phone to skip that conversation: The worries and the deaths I pulled a shift in ICU while I was at Papworth, and I attended a cardiac arrest call (someone's heart had stopped). It was on a COVID ward, so I PPE'd up and turned up. It was a man who was around the same age as my dad and also from a BAME background. I don't normally notice a patient's skin colour, but it was very apparent that there were a lot of BAME people on this ICU, especially since Cambridge isn't known for a high BAME population. They were doing chest compressions, and the blood gases had just come back. pH of 6.9 (normal 7.35-7.45), pO2 of 5kPa (normal 11-14kPa). Not compatible with life. The X ray was just completely white where the lungs should be. He had such severe COVID that even on full ventilatory support and every drug imaginable, he wasn't able to get much oxygen into his blood through his lungs. He was so hypoxic that his heart had stopped. They switch the person doing chest compressions routinely to make sure the person doing them doesn't get tired. They asked me to take over. It's the first time I've done chest compressions outside of an ATLS course on a real person. A real person is warm. They have ribs that had been broken from the force of the compressions. They had a heart that I was suddenly pumping for them manually. His heartbeat came back, but we checked his eyes with a flashlight. Not responsive to light and dilated. He was brain dead. Suddenly all the ethics and law lectures about DNACPR came back to me. I remmeber thinking at the time that how could we ever just give up and not even try to resuscitate someone? Why would we not give it our all? Suddenly it all made sense. I felt like I'd really grown up very quickly, that my perspective on it had changed. Here he was, back "alive" but not quite the same. I listened in to the call to the family. They were desperate and bargaining and lashing out, "how could you say he's giving up? he's a fighter and he's so young, he's so healthy, he can make it through this! Why can't you put him on ECMO or something? You're not even trying!". But they had last seen him in last March. They had no idea about what was happening right there and they couldn't even come to hold his hand as he was passing away. He was peri-arrest, and his heart stopped again about ten minutes later. His lungs had not been recovering at all, and all the machines, the ECMO, were only there to support him as his body healed. It never did. All of it was quite a harrowing experience, and I think the team of doctors, nurses and other healthcare workers handled everything the best they could. I can't think of any better way of managing the situation or a better way of breaking the bad news to the family. But it's times like that I felt quite powerless, that there was nothing that could be done. I'm okay, although it would be weird if I didn't feel weird about the situation afterwards. I think on reflection, it's made it even clearer that medicine is the right career for me. It's those experiences that I remember when I'm studying, to remind me why I'm learning all of this stuff in the first place. 2020-11-19T00:00:00-00:00 https://souradip.com/blog/interview-simulator I've always found that the best way to practice for an interview is to record myself answering questions, playing it back, then reflecting on how I could improve. I recently did some work with a charity for widening access to studying medicine and made them an interview simulator that allows you to do exactly this! We Are Medics @WeAreMedics_We’ve filmed 5 full, realistic and INTERACTIVE medicine mock interviews for you to complete anytime 😍These are exactly like the real thing will be in 2020, and your responses are recorded so you can watch them back + use our mark scheme to reflect 😌www.wearemedics.com/interview/14 07:43 PM · Nov 7, 2020 In it, myself and some friends are asking questions, you record a response, and continue as if it was a real video-call interview! Then once, you've recorded a response to all of the questions, it plays back your response and you can mark yourself with the built-in mark schemes. Have fun! 2020-06-19T00:00:00-00:00 https://souradip.com/blog/neurotransmitters-not-like-filling-up-a-cup In the wider community (and in medicine) I often see drugs that affect the brain be misunderstood. A popular idea seems to be that our brain is a cup full of some amount of serotonin, when we are depressed it means we have less of it and that when we take an SSRI, we increase how much we have in our cup. This is very simplistic but it's misleading. We might not understand everything about the brain yet, but we can certainly do a better job than this. Imagine we are talking about computers instead of brains. We notice that opening Microsoft Word uses more electricity. But computers are more about just the amount of electricity they use or not. It's where they are what what they do. If we fill the computer with more electricity, it won't make more Microsoft Words open. Neurotransmitters are like our brain's form of electricity, because they help conduct specific messages along nerve "wires". How does a synapse work? Okay, what do these drugs specifically do? First let's see how a neuronal synapse works, the basic building block of our brains: Electrical impulse comes in Neurotransmitter released Neurotransmitter stimulates a signal on the next neuron Neurotransmitter taken back up so the signal doesn't stay on forever But it's important to note that our brains aren't just one massive synapse. Our brains are complicated circuits, and different parts of our brain use different neurotransmitters to perform logic and do calculations. Some nerves use different neurotransmitters for different types of messages. Let's take the example of methylphenidate (Ritalin), often used to treat ADHD. This is a dopamine reuptake inhibitor. When a dopamine signal comes in, it will linger for longer in the presence of Ritalin because the drug is blocking the removal of the neurotransmitter. This will then fire a stronger signal in the next neuron than what would be triggered otherwise. In other words, it will amplify any signal that uses dopamine. What is attention? What is attention? Broadbent proposed a "filter" model in 1958 - attention acts to filter out other information to prevent processing overload in our brains. We know that reward is mediated by dopaminergic neurons in the nucleus accumbens in our basal ganglia, a small but important part of our brain. Many experiments have shown that a dopamine signal is created here when what we expect is different to what we have seen, a bit like an "error signal". The "error" is the fact that something unexpected happened. We then adjust our mental model and try again, and we get less of an "error" signal. (Exciting work by Wolfram Schultz and co!) This is why we get less and less pleasure out of doing the same thing again and again, but at the beginning things seem exciting and new. We can become addicted to almost anything, as long as whatever we are doing produces a reward/"error signal" that hits escape velocity, and somehow bypasses the default tendancy for something to get less and less rewarding the more we do it. This could be through self-administering recreational drugs (interestingly, addiction doesn't tend to happen so much when "addictive" drugs are given in a healthcare setting by someone else, because it breaks this feedback loop), or even finding video games so fun that they just don't get boring. This also seems key to maintaining our focus and attention on something. If we aren't getting a rewarding feedback, it's like getting told to study and that there's a test next week, vs being told to study but there's no test. A lot fewer people will have the motivation to study. How does Ritalin affect attention-deficit? So if we combine the two bits of information above, taking Ritalin will amplify any dopaminergic signal that our brains are producing already. It will specifically affect the bits of wiring in our brain that use this as a messenger, and make them shout louder when they have something to say. It will not create any new dopamine, or any new signals, but if we assume someone with ADHD has a problem with attention, then they will get more of a reward out of doing something than they were before. This will hopefully get them to focus on things to the same degree as everyone else. Extending this more nuanced view to other conditions We can extend this view to other conditions, such as depression. Taking an SSRI won't turn up the amount of serotonin in your brain. Instead, by preventing their reuptake, it will amplify the signal specifically in the nerves that use it as a transmitter. Over time, it will mean the emotional signal that your brain uses to form memories (serotoninergic neurons project from the raphe nucleus to the hippocampus, which is important in memory formation) will be less negative. When you record new memories (from your waking experiences or even through the process of remembering memories you already have, a process that actually destroys and rewrites them [but which is a whole topic for another day]), it's likely that process will integrate some of this more positive signal to colour your feelings of those memories (prosody). Eventually, over a period of months, you'll notice a gradual mood change because the memories you'll draw from will be more positive. Conclusions Overall, the brain is still very mysterious and we do not understand a lot of how it works. Drugs that affect the brain are blunt tools, but they are cleverer than just turning up and down the amounts of neurotransmitters everywhere in your brain. It's how they amplify specific types of signals inside certain circuits in your brain that can help us understand both why they work and ask questions to what's really going on inside our heads and make us who we are. 2020-06-17T00:00:00-00:00 https://souradip.com/blog/nhs-bursary-guide-for-medical-students I am about to return to Clinical School. The NHS Bursary Scheme pays for tuition in our fifth and sixth years of medical school, and it's the time where we need to apply to be ready for the coming academic year. The information to make a successful application, however, is scattered all over the place! I had to call their support line on a few different occasions (they are very helpful - 0300 330 1345!). Gemma Wren @gnwrenI don't know why people complain about the NHS bursary application process. I mean, all I have to do is find my original birth certificate (which was lost in France in 1996), post them my passport, my parents' decree nisi, and a vial of my own blood collected during a full moon.359 08:01 PM · May 29, 2020 I thought it would be helpful for anyone applying in the future to have all the information in one place, from submitting the application online to sending in evidence (and the specific type of Royal Mail envelopes you'll need) through to how long you can expect it to take. The timeframe There is a suggested timescale posted at NHSBSA - but they advised me to start earlier rather than later because posting evidence to them and getting it back can take a while. They recommended I start even if the "application window" hasn't opened just yet. For this year: Date your course begins Applications open Application deadline June / July 2020 2 March 2020 29 May 2020 August 2020 6 April 2020 26 June 2020 September / October 2020 4 May 2020 31 July 2020 The Online Application This is probably where there is the most guidance available. They have a step-by-step guide over here. You can manage this all online through BOSS. For my situation (if you're an MB/PhD student, this gets somewhat confusing) I had to fill out my form for course details as follows: Filling out income as a medical student with a part-time job For the income section, they state If you are a full-time student, you should exclude earnings for work done in the evenings, at weekends or during holidays whilst you are attending your course. Only declare these earnings if you are a part-time student. So if you're a full-time medical student like me, even if you have a job, everything here should be 0. All expenses should also be filled out as 0: If you are not declaring any income in this section, please do not declare any expenses e.g. your rent. This is because we do not take into account your expenses when you have no income to offset against them. This will also save you time, as we will not need you to send us evidence of this e.g. a rent agreement. Financial independence status for MB/PhDs For the Student Award section, MB/PhDs do not sadly fit the criteria for being an "independent student" by default (despite being financially independent from my parents from a research grant for 3 years) and thus eligible for means-testing based on our current income. The criteria for this is you have supported yourself out of your own earnings (including any benefits) before the first academic year of your course for a total of 36 months? which is likely not the case since the start of my course was when I was 18. There are however exceptions if you meet any of the other criteria, e.g. if you are married/divorced/widowed, parents are deceased/irreconcilably estranged, if you have a child, etc, none of which I have thankfully managed yet by 25! Posting evidence I was not applying for the means-tested portion, only the tuition fees. Even so, I was requested to submit evidence of my identity and proof I am on a medical course: Please provide your most recent local authority (LA) or student finance award letter. Please provide two forms of identification. This must include your birth certificate AND one form of the following photo identity: Passport (current and valid) Driving Licence (photo card) So I had to send in my original birth certificate and passport, as well as my latest student finance letter AND their student coversheet from my application. These are not things I want to get lost in the post. They recommend: Please send your NHS Bursary supporting evidence and student coversheet by Special Delivery post to: (Your unique SBA reference number) NHS Student Bursaries Ridgway House Northgate Close Middlebrook Horwich Bolton BL6 6PQ Special Delivery post will enable you to track the delivery of your documents. We will email you once we have scanned your documents but please note that this may be up to five working days after we receive them. We do not offer appointments or accept walk ins to have your evidence scanned (any walk ins we do receive will be turned away) so post your supporting evidence to us at the address above. But my local post office was adamant you cannot "pre-pay" for special delivery. They were happy to give me these grey envelopes for free though, to be paid once it's ready to be posted, price dependant on weight. To get around this, you need to ask for Special Delivery stamps x2 (which are valid for envelopes <500g) as well as the grey special delivery envelopes x2. Yes they are cheap thin plastic. You can ask for A4 envelopes which are a better fit for paper than the A5 ones they have. Fill out one envelope with your home address (that they will send it back in) and take a photo of the barcode sticker on the envelope (e.g. SF847997909GB) - this is the tracking code for your return envelope that you can use at Royal Mail Track and Trace. Then you must fill out the other envelope with the SBA number found at the top of BOSS after logging in followed by the address above. Take a note of the barcode sticker on this too (the SF....GB number) to track its progress to NHSBSA. Finally, put the envelope with your return address on it inside of the envelope to send to them, along with the rest of the evidence they request. Then tape it shut. If you do not provide a return envelope, they said that they will post it back by standard second class mail. Wait to hear back You can track when they receive your evidence, but this may take up to 15 working days when I called them up. Nothing will update on BOSS until they have processed it all. My evidence arrived on 2nd June, heard nothing after that. I called them up to make sure they had in fact received my package, and they confirmed receipt over phone. I got an email saying my application had been approved on 17th June and my passport/birth certificate/etc arrived back on the very same day. Fill out BOSS again, this time with bank details Now that the application has been approved, you will need to log into BOSS again, and this time go to "Update Bank Details" to fill out where they should pay the non-means-tested (and means-tested if applicable) grant to. You may also need to contact your medical school to submit a proof of enrolment: Dear Souradip, Thank you for your application for an NHS Bursary. I have calculated and approved your bursary award, however, I am unable to release payments until we receive official written confirmation from your university that you are eligible for an NHS bursary. This is known as a BUR99 (confirmation of eligibility) form If you have now started back onto your course, please check with your university that they have arranged to send a BUR99 to us. Conclusion It takes persistence to make it this far and can't be left to the last minute. It is well-worth it though, to be able to finally stop adding to the £60k+ debt already incurred through student finance for our medical degrees! 2020-06-14T00:00:00-00:00 https://souradip.com/blog/books-vs-internet Books changed the way we think about thinking. The printing press changed the way we think. Knowledge was no longer restricted to just experts, it could be spread to anyone. And just like Isaac Newton, we could all now stand on the shoulders of the knowledge gained from others. It was possible to think more creatively because we could now break down the silos of different minds and come up with new insights nobody could have come up with alone. I'm going to specifically talk about non-fiction as "books" here, but many of the similar points apply to fiction too. The internet has made knowledge far more accessible than before 👍🏽 The internet has changed the world of knowledge completely. Imagine a time before you could look up the answers to your questions at the click of a button on Wikipedia or Google! More than that, access to information is now 24/7. To answer that niche trivia question you no longer need to be dedicated enough to go to a library, scour through book after book until you find something you're looking for. Research was made easier, you no longer needed to dig through every journal to be on top of the recent developments in the field. You can find the latest, up-to-date papers just with a few keystrokes. Literature reviews become much less labourious. The internet has AN answer for everything? 👎🏽 The problem with the internet is that it only has ONE answer for everything, or far too many to be able to sort out the garbage. Wikipedia will have an article, but it is one person's understanding. This is systemic and intentional; you cannot create a secondary article on any topic that exists. And most editors are not willing to let an existing page get a complete re-write unless it's factually incorrect. You can get many different books about a certain topic. If you read one and you don't quite understand what they're on about, you can just find another one and try your luck again. This is why we have lectures, small class supervisions, essay writing at University. The more ways we are exposed to some information the more likely it is we'll "get" it. WebMD is a joke for telling you every little symptom you have is cancer. Wikipedia will have the very specific details of the chemical reactions needed to produce a drug, but no discussion on how it works and how it relates to a big-picture view of human physiology. It is notoriously bad for anything mathematical; everything is explained to the level of a Maths PhD without explaining any of the reasoning that led to it. Just the end result. This robs us of any critical thought or ability to assess the process of how the information came to be. What's bad on Wikipedia is even worse elsewhere. The internet has a notorious problem with verifying information. If you stray too far from the one or two authorative sites, you'll likely end up in the lands of conspiracy theories and "alternative facts". Books solve this problem by firstly having a barrier to entry, so that publishers need to decide something is worth it before printing it, and secondly by being possible to use the internet to get reviews on books so that you know you're getting a good one, something that's impossible to do with websites. Books provide continuity of thought 👍🏽 This is key whenever the piece of information is not a short snippet you are after. If I want to learn the piano, I could watch YouTube videos all day long, but each video is a short snippet of the process of learning. After I've done that one, I don't know enough to be able to find the one that is slightly harder that I can still do. They aren't ordered to make it easy for me to learn. On the internet, there's just a web of loosely related information, algorithmically generated to maximise my viewing time. A book will have a series of exercises that have been carefully thought out to maximise learning. It will have a structure that goes in order of difficulty, and the confidence that it has worked successfully for many people. You can't say that about a wikipedia entry. Continuity of thought is important because it lets us logically reason with the information we are given. Just an answer doesn't do that. The process is more important than the destination. The internet only has the destination. Conclusion The internet is good for superficial knowledge about lots of topics, books are good for in-depth knowledge about a single topic Books have evolved through literally hundreds of years of evolution to distill topics into a format that is easy to learn from. The first few decades of the printing press probably weren't very good compared to now. The internet will probably improve as we find new ways of addressing these issues. But for now, I think books are still the best way to learn about a new topic in-depth, while the internet is a good way to learn about lots of topics superficially. If you want to learn about something new, or a new skill, the best way I think is to find the best book about it online, and then go read the book instead. 2020-06-13T00:00:00-00:00 https://souradip.com/blog/on-the-failure-of-complex-systems Complex systems don't fail just because of one reason I once ran to catch a train back home from University. I ran and ran as fast as I could. I got to the platform mere seconds after it had closed its doors and started to pull away. I had been so close, but I'd missed it. What had been the culprit? What could I blame? Was it the old lady in front of me in the queue taking her sweet time with her coins while my train was waiting for me? Was it the dodgy internet connection along the way that hadn't let me buy a ticket on the way there? Had I not ran fast enough? Did I wake up too late? Did I spend too long helping a friend through an emotional crisis the previous night? Missing a train is a complex system. In truth, the real reason was all of the above and yet none at the same time. Complex systems have a lot of redundancy built in. They can handle one or two things not being quite perfect. But when enough things go wrong, they all have an effect that's bigger than any one of them combined. The Missing Heritability Problem In genetics, there's this curious problem called the "missing heritability problem". We know that, for example, 60% of the incidence of schizophrenia is due to genetics through studies of twins. But the genes we've identified only acount for a small fraction of that. In the 2000s, the Wellcome Trust used genome sequencing to try and find genes responsible on a massive scale for common diseases like coronary artery disease, Crohn's, Rheumatoid Arthritis, and more. But again, they only found a few individual genes that were associated with these diseases. The human body is a complex system We are full of redundant systems within ourselves. We've evolved over millions of years so that we can recover from any single hit to our health. The analysis that's been done on the genome sequence studies have focused on assessing individual gene associations with a disease, or polygenic risk scores from adding together the individual risks from each individual gene. I think it's likely that the missing heritability comes from somewhere else. The individual risk of me missing my train from each of those factors I mentioned earlier are minimal, even added together. But they have a synergistic, superadditive effect when they all happen together. I think a similar model can be used to suggest that many of the genes responsible for these diseases are hiding, only to be brought out when the right context of simultaneous mutations exist. Unexplored avenues of research I think this represents a fascinating unexplored avenue of research. With the advent of neural network architectures like transformers, originally developed for machine language translation, I think this could be feasible to tackle now. The translation for any work to another language requires knowledge of the context of the surrounding words in the sentence for it to be appropriate. These neural networks allow not only to infer the combinatorial superadditive effects that traditional statistical techniques cannot, but we could then run the model backwards to see given a certain gene, what other genes are taken into consideration when deciding if someone is at risk. We even know the theoretical maximum accuracy for these models (the heritability), beyond which it would be impossible to improve upon. We could even build useful clinical prognostic tests based around these, rather than just saying someone has a "4% increased risk of X" as sites like 23andme seem to do these days. 2020-06-12T00:00:00-00:00 https://souradip.com/blog/type-ii-diabetes-a-tragedy-of-the-commons I'd like to describe a few of my own thoughts on how type II diabetes develops. By taking a step back and looking at a "big picture" overview of how this disease starts, perhaps this might help us make sense of it in a way that looking at too small a scale cannot. Maybe this disease is not the result of something going wrong, but is the result of all the cellular machinery working exactly as intended, and then failing in an unexpected way when all of the parts fit together as a system. The tragedy of the commons is "where individual users, acting independently according to their own best interests, behave contrary to the common good of all users by depleting or spoiling the shared resource through their collective action." What is type 2 diabetes? Type 2 Diabetes is a chronic condition is best characterised by high blood sugar level (above 7 mmol/L) and insulin resistance that then goes on to cause cardiovascular disease, retinal disease, etc through the formation of advanced glycation end-products. (read more here) Usually, when we have a high blood sugar level (e.g. after eating a sweet!), our pancreas releases insulin. Insulin acts on our tissues to signal them to take up glucose into the cells away from the blood, thus bringing the blood glucose level back to normal. In type 2 diabetes, insulin resistance means that the insulin that is normally released has less of an effect than in a normal, healthy individual. This means cells take up less glucose out of the blood, and so the blood sugar level stays elevated, and given enough time, this can result in all of the damage to our bodies mentioned above. With the defining feature of insulin resistance, many people focus on the "insulin" part as the bit that is going wrong and causing disease, and we often use insulin to treat type 2 diabetes. From the NHS website: It's caused by problems with a chemical in the body (hormone) called insulin. It's often linked to being overweight or inactive, or having a family history of type 2 diabetes. While technically correct, this kind of definition makes it sound like the insulin is doing something wrong, and if we could fix that then the disease would go away. And it makes it sound like if you have a family history of it, you are at risk whatever you do and it is not really under your control. I think it's slightly more subtle than that. Insulin is a signal for cells to take up glucose and store it What does insulin do to get cells to take up glucose out of the blood? Here's a molecular signalling diagram explaining how: Broadly, the sugar has to go somewhere. If it's not being immediately used (the most likely scenario, we tend not to exercise and eat at the same time!) then it is stored as glycogen. What happens when the cellular energy stores get full? What if this stored energy is never used? This is often the case in modern day society. There has been no evolutionary reason to select for tolerance to this, as we have often lived in times of starvation in the past, not abundance. Physically, a cell must only be able to store so much energy, beyond which it will no longer function. Perhaps there is a cellular stored energy sensor that eventually goes "hold up, there is no more room, stop taking up energy when insulin tells us to". In type II diabetes, we see cells downregulate their insulin receptors and this means they become resistant and refuse to take up glucose. ** But what if this isn't a result of the disease causing them to downregulate these receptors, but simply a normal mechanism because they can't take up any more energy? ** What happens when too many cells refuse to take up glucose because they are too "full"? One or two cells doing this would not be an issue, as we have plenty of cells that are able to take up glucose. But what if after many many years we reach a critical mass of cells such that our blood glucose level slowly starts to slide out of control? This would be the classic picture of type II diabetes as seen in endocrine clinics in hospitals. This explains many of the features of how type II diabetes develops We often think of type II diabetes as one associated with obesity, somewhat intrinsically linked as part of a "metabolic syndrome". I speculate that this large scale systems-based thinking of these components might help us make more sense of this cloud of symptoms and associated conditions. Genes that predispose to the common forms of type II diabetes as discovered by GWAS studies have been largely to do with the brain. Specifically, appetite control. It is well known that given a high energy food environment, different individuals will respond differently, and a lot of it is thought to be down to how our brains are wired. This is why it is an acquired condition rather than congenital, and largely affects people after a certain age once all of that excess energy has accumulated. Also raises some interesting hypotheses - could people who undergo liposuction surgery be more at risk because they've lost their reserve of fat cells? This model also explains how type II diabetes often develops very gradually (hence the term pre-diabetes). Our bodies have a lot of redundancy built in - as individual cells become "full" and stop responding to insulin, there are plenty of other cells that can pick up the slack. It is only when a critical mass of these cells are unable to take any more up that we tip beyond the threshold of a safe blood glucose range. Many treatments affect some part of this pathway and explains their unintended consequences This gives us a framework to think through the possible consequences of interfering with individual parts of this pathway, and whether it would truly be beneficial or not. For example, some drugs (like metformin) act by fooling the cellular energy stores to take up more (thereby "decreasing insulin resistance"), while others (e.g. recombinant insulin) cause a higher activation in the insulin receptors left on the surface of these cells, stimulating them to take up more. These drugs also have the side-effect of causing further weight gain - perhaps this is a direct consequence of its mechanism of action rather than a "side effect" as we often describe it, followed by resistance to treatment once patients reach their new "limit" of cellular energy stores. We can use this model to generate new hypotheses What implications does this have for how we treat this common condition? Do many of our front line therapies, by acting to fool these metabolic sensors actually make the problem worse by unleashing the consequences of overburdened adipose tissue and subsequent inflammation as they inevitably burst open? If someone congenitally is born with less adipose tissue, is this the reason some very thin people develop type II diabetes, because they reached that critical mass of cells full of energy at a healthy BMI? Does this explain why some very obese people live for years without suffering from obesity, because they had so many more adipocytes able to absorb all that energy? I propose this might help us make sense of the often confusing data we get, and to put all of these findings into a framework, to turn all the associations between variables into a coherent story with causal relationships between each of the parts. From the same NHS website, It's a lifelong condition that can affect your everyday life. You may need to change your diet, take medicines and have regular check-ups. Is it necessary for type 2 diabetes to be a lifelong condition? Or is it possible to change that perception by showing that it is simply the result of the body's energy stores being full, and that by emptying these stores it is possible to no longer have this condition? Various published case studies suggest so! Conclusions This is just a thought experiment based on my experience as a medical student on wards, lecture notes and literature review. I am open to new ideas, this has just been my interesting take on the "big picture" story of diabetes. 2020-06-11T00:00:00-00:00 https://souradip.com/blog/comparison-the-thief-of-joy This is a quote often attributed to Theodore Roosevelt. During this time in lockdown, I have had a lot of time to mull it over. I've come to the conclusion lately that the only comparisons that we should be making is with ourselves. This has been the key to my quest for self improvement. The more I compare myself to others, the more I risk hurting my motivation to ever get started. Let's take running. I've never run much before in my life. I thought I might try getting into it with Strava now that I have all this time. The easiest thing to do is to check out the segments, see what other people have managed to do, try and outcompete these olympic athletes with superhuman times inevitably to be disappointed, and not keep at it. Comparing yourself with others is a one-way road to misery and insecurity. The key to self improvement is consistency. Nobody ever gets good at something overnight, not repeatably. But keeping at something while being humble is a great way to get better at just about anything. I have no dreams of competing with Usain Bolt. I just want to live healthier. And the key to consistency is motivating ourselves enough to keep at something. In those two things lie the secret to improvement - self-comparison. I am running my own race, against me from the past. With a week's worth of practice I am better at this than me from a week ago. Even though it happens gradually, I can measure this improvement, and the data does not lie. So I started. I couldn't run a hundred metres without getting out of breath. That's okay, I noted that down. A few days later, my pace was faster. Within a month, I was running a whole 5k without stopping once. Within two months, I was doing a 10k every day, with a 15k sprinkled in for good measure. I'm better than the me I was yesterday, and that is enough. There is no grand leaderboard of winners at life at the end of the day, the key to happiness I've discovered has been to be happy with what I've got, and to be happy with what I've won through commitment to a cause. 2020-06-10T00:00:00-00:00 https://souradip.com/blog/arthur-max-barrett-a-cambridge-pathologist When I was a third year undergraduate at Cambridge, I did some research during my Part II course in Pathology (looking at the genetic causes of primordial dwarfism). We discovered a new mechanism that a certain gene mutation (in Cenpj, part of the centrosome) causes this form of dwarfism, with implications for how our bodies limit our size, and perhaps how cancers bypass this to keep growing indefinitely. I presented my thesis and the department awarded me the "Max Barrett prize for Best Project in Cancer & Genetic Disease". But who was Max Barrett? TL;DR He's the father of Syd Barrett from Pink Floyd, the rock'n'roll band, and related to Elizabeth Garrett Anderson, suffragette and first female doctor to qualify in the UK. Early years Arthur Max Barrett, MD (28 July 1909 – 11 December 1961) was born in Thaxted, Essex, but moved to Cambridge in his teenage years. He went on to study medicine at Pembroke College, Cambridge in 1927 having won a State Scholarship. There he obtained a considerable number of awards and honours: a Major Scholarship in 1928; a Schoolbred Scholarship in 1930; a First Class place in the Natural Science Tripos Part I in 1930 and in Part II in 1931. This Part II course in Pathology (the same one I studied!) was introduced in 1925 by Prof. Henry Roy Dean, who he was very much influenced by. This was followed by a Foundress Scholarship in 1931; five prizes during his clinical training in the London Hospital Medical College (where he went as an entrance scholar in Pathology). He graduated MB BCh in 1934. A renowned teacher He worked in the wards and laboratories of the London Hospital from 1934 to 1938 and was University Demonstrator in Cambridge from 1938 to 1946, the only one in the Department of Pathology during the war years. Even during this period he managed somehow to continue original work on the Paul-Bunnell test, as well as introducing his valuable picric acid method for removing “formalin fixation pigment” from sections. His research His research, which had commenced in the Clinical Laboratory, was pursued vigorously at the Bernhard Baron Institute where he held a Halley Stewart Research Fellowship. His observations drew him to the study of “target” corpuscles. He found that these misshapen red cells are in fact “bowl-shaped” in fresh blood and that they assume the form of “target cells” only in fixed films; he initiated the concept that the special resistance of the bowl-shaped red cells to haemolysis in hypotonic saline is proportional to their large relative surface area. When he returned to Cambridge in 1938, as well as teaching he was actively interested both in the routine pathology services of Addenbrooke's Hospital. In 1946, when those services were saddled to the university, he became consultant for the hospital as University Morbid Anatomist and Histologist. He became a very prominent pathologist. N.B. He is different to Norman Barrett, for whom the eponymous Barrett's oesophagus is named. His research spanned many different topics, from developing a test for infectious mononucleosis (glandular fever) to determining the cause of toxicity from mustard gas. He is said to have believed in the precise use of language and attention to detail, and often able to discover some small but important clue that would lead to a firm diagnosis. He is remembered in Addenbrooke's by the Barrett Room, named after him. His later years During his later years as University Morbid Anatomist he somehow found time to carry out a brilliant investigation on arterial hypertrophy. This work was embodied in his thesis for the degree of M.D., for which he was awarded the Raymond Horton Smith Prize of the University. He noted that Turnbull had believed it possible to form an approximate estimate of the increase in weight of the heart by examination of the arteries; but Max stated with typical modesty that for observers “less experienced than Turnbull estimations based merely on inspection are likely to be unreliable”. Barrett’s 'histological eye' indeed was among the most gifted and experienced of his time, yet it was characteristic of his native honesty and meticulous accuracy to demand a method of assessing arterial hypertrophy that is free from subjective error. For this he developed a highly efficient quantitative method of assessing “relative thickness” and cross-sectional “relative area” of the tunica media. By use of his “undulation index” Barrett took full account of the degree of post-mortem contraction of arteries, a factor which had vitiated so many previous investigations. Barrett’s solution of this problem is a notable advance in angiology, and will always be a basic model for those interested in transferring histological observations from the art of opinion and impression into the exact science of quantitative measurement. His legacy He was also a member of the Cambridge Philharmonic Society as well, where he was Honorary Secretary for more than 20 years, and this musical legacy is something that he imparted on all of his children. It is reported that they had many impromptu family music nights. He passed away from cancer while his son Syd, who would go on to start Pink Floyd was 14, and it is quoted as being the first major life event that would go on to contribute to his later mental health issues and struggles with LSD. Suddenly, the Barrett family's world was turned on its head. Inoperable cancer was diagnosed and Max Barrett died suddenly on December 11. Syd had religiously kept a diary ever since his eleventh birthday. He left the entry for this day blank. 2020-06-09T00:00:00-00:00 https://souradip.com/blog/modern-day-alchemy-in-vitro-platelets Originally posted at the NHSBT R&D Blog. What does a baby, the atom bomb and Dolly the Sheep have in common? Stem cells! We all start out life as a single cell, a fertilised egg. Somehow, that cell divides and divides and its progeny form every single one of the approximately 30 trillion cells in our bodies, from skin and bone through to our hearts and brains. Throughout much of the Cold War, many people worried about the effects of radiation poisoning. Radiation affects cells that divide quickly, especially blood - red cells only last about 120 days in circulation, and platelets only 5 days. Often people died within weeks of bone marrow failure; they were not able to make any new blood. Research found that by transplanting just one specific bone marrow cell, it was possible to regenerate the entire blood forming system. This special cell is a blood stem cell and we now routinely use radiation and bone marrow transplantation to cure many blood cancers. It also used to be thought that once a stem cell differentiates, it can never go back. Skin makes more skin; it doesn’t turn into muscle or blood! But in 1997, Campbell and colleagues took a fully differentiated skin cell from a sheep, took its nucleus, and used it to create a complete sheep (Dolly)! This paved the way for the idea that given the right factors, our own differentiated cells can have their internal clocks turned back, to turn them back into a stem cell just like the one that makes all of those trillion cells that make us who we are, and one day we might be able to use them to grow tissues for transplant. Yamanaka and colleagues discovered in the mid-2000s how to “reprogram” fully differentiated cells (like a bit of your skin) back into a stem cell state, winning them a Nobel Prize – these are called induced pluripotent stem cells, or iPSCs. Unlike most things we are used to in life, these stem cells don’t run out – they can and do make more of themselves; all they need is the right care, attention and nutrients! Platelet transfusion is a life-saving treatment and demand is increasing Up to 2.9 million transfusions of platelets take place each year in Europe. Platelet transfusion strategies are driven by either the need to stop active bleeding (therapeutic) or to prevent its occurrence in at-risk groups (prophylactic). The most common situations are for patients receiving chemotherapy, requiring surgery (especially cardiac surgery) or having experienced severe haemorrhage (e.g. from pregnancy or childbirth, trauma, largely from road traffic accidents, or severe anaemia in the young, often caused by malaria), often being combined with other blood products such as fresh frozen plasma (FFP) and red blood cells (RBCs) in a massive transfusion protocol. In most high-income countries, there is an adequate supply of blood with its use largely pre-planned and predictable, with most blood transfusions (79%) taking place in the over 60 group. Blood donors and their donated blood are rigorously screened, and so the frequency of disease transmission through transfusion is low, although tragically marred by recent scandals in the 1980s when strict screening and testing for infectious agents was less common. However, there are major ongoing issues in lower-income countries, where most blood transfusions (67%) are given to children below the age of 5, reflecting the different demands on medical care in different parts of the world. In sub-Saharan Africa, blood shortages (the leading cause of maternal mortality due to post-partum haemorrhage) and unsafe blood (leading to many instances of transmission of HIV and hepatitis) represent major challenges in transfusion medicine. These are thought to be due to the result of a relative scarcity of donors (combined with a free market, quickly leading to blood products becoming unaffordable for many), the unwillingness of relatives to donate due to cultural differences as well as inadequate supply chains, storage and transport infrastructure. Figure 1. Demand for platelet units increasing over time in the UK, adapted from (Cowan, 2017). The demand for platelet products is steadily rising due to their increased use in chemotherapy and trauma surgery. At the same time, the supply from donors has remained largely constant. There have been no major advances to improve the storage of platelets beyond 7 days, compared with 35-40 days for red blood cells, making these a comparatively perishable blood product. Due to their short shelf life, supply and demand is carefully managed, especially over holiday periods when fewer donors are available, and after natural events such as snowstorms. Stem cells offer us a way to get blood without relying on blood donors In 2016 our lab published a new method that allows us to take iPSCs and push them to turn into platelets. This means we have been able to make blood from a renewable source of stem cells on demand! We do this by activating three transcription factors, GATA1, FLI1 and TAL1 inside of these stem cells, which forces them to divide to make megakaryocytes, the precursor cells to platelets. We can then put them into a special bioreactor that holds these cells in place while they give off platelets, just like they do in our bone marrow! Figure 2. Overexpressing just three transcription factors pushes iPSCs to differentiate into megakaryocytes, the precursor of platelets. Diagram adapted from (Moreau et al, 2016) Since then a lot has happened, and many labs around the world are racing against each other to be able to scale up production of platelets like this. One platelet transfusion unit contains 3x1011 platelets! Some labs (including our own) have even modified these stem cells using cutting-edge genome editing technologies (like CRISPR, TALENs, Zinc Finger nucleases) to remove the molecules on the surface of normal platelets to make them immunocompatible with everyone – a bit like having on-demand O-negative universal platelets! My PhD has focused on engineering these lab-grown platelets as a drug delivery system and engineering them to contain clotting factors to make them more potent and allow us to do more with less! So far, we’ve only ever looked at the effect of these in vitro derived platelets in animal models, such as mice. Figure 3. The Ghevaert Lab, November 2019. L-R: Rebecca McDonald, Amber Philp, James Warland, Moyra Lawrence, Daniel Howard, Winnie Lau, Souradip Mookerjee, Amie Waller, Adam Pullen, Cedric Ghevaert, Holly Foster, Amanda Evans As we get closer to the point where human clinical trials are becoming feasible, it’s more important than ever that we all have the same ways of assessing the purity, safety and function of these in vitro platelets compared to donor platelets, so that we can all compare the progress we are making and make sure what we do will not do harm. The World Health Organisation (WHO) have recently commissioned a standard for assessing these next-generation medical products. I recently published an article in the journal Platelets about how far this field has come, proposing a few of the standards we use to assess our platelets, and made note the challenges we still face. It’s our collective responsibility to leave the world a better place than how we found it, and stem cell research offers us one way to that brighter future! 2020-02-07T00:00:00-00:00 https://souradip.com/blog/travel-report-galveston-tx Originally posted at the British Society for Haematology website. "I had the opportunity to discuss and receive valuable feedback on my work from many different perspectives, which would otherwise have been impossible". I was grateful to receive the BSH Travel Grant to be able to travel to the Gordon Research Conference on Megakaryocytes and Platelets, in Galveston, Texas, USA. I am an MB/PhD student currently studying for a PhD in the lab of Dr Cedric Ghevaert at the Department of Haematology, University of Cambridge, and the lab works on generating platelets in vitro for transfusions without the need for blood donors. My project in particular focuses on developing engineered platelets as a targeted drug delivery system, by loading the granules of platelets with drugs, so that they can activate and secrete these drugs where they need to go, without systemic side-effects, and thanks to the help of the BSH I was able to present my work at this international meeting, where I was able to meet the authors of all the big papers in the field of platelets biology! I had the opportunity to discuss and receive valuable feedback on my work from many different perspectives, which would otherwise have been impossible. I also had the chance to see my research in context of the other fascinating advances in the field taking place. The conference was very well organised, with several sessions each with a different theme. Each session started with short introductory slides presented by session chairs, followed by several presentations. I was also able to talk to my contemporaries working around the world on the similar issues we face, as well as the big names of the field and felt I was able to learn from both and made lots of new friends along the way! One of the reasons for attending the conference was to find out more about the tools that other labs use to ask similar kinds of questions that we are. One of the most fascinating talks was about some of these tools from the Hahn lab at UNC Chapel Hill. They had developed biosensors to detect signalling GTPases spatiotemporally and study how the dynamics of signalling take place in much more depth than can be shown by any western blot, developed computational techniques to analyse these videos and even engineered optogenetically activatable GTPases, using this to demonstrate how they were able to activate Rac61 precisely and reversibly using blue lasers. Rac61 is thought to control cell motility, and they were able to show how localised Rac activity was enough to mobilise cells to follow around the laser pointer, much like a cat. 1 From this I learnt a lot about the frontiers of what was possible with protein engineering as well as about new techniques for studying these signalling cascades in a way I had not considered before and certainly plan to make use of going forward. Overall, I would like to thank the BSH for this opportunity to be able to travel, present my work, discover more about the work of others and the challenges we are facing together, and foster potential collaborations in the future that will hopefully benefit us all. Wu, Y. I. et al. A genetically encoded photoactivatable Rac controls the motility of living cells. Nature 461, 104 (2009). 2019-11-21T00:00:00-00:00 https://souradip.com/blog/mb-phd-symposium Being the final year of my PhD, I was asked to present a 20 minute talk at this year's annual MB/PhD Symposium here in Cambridge. The talk was very well received! It was also a special year, being the 20th anniversary of the Cambridge MB/PhD Programme. In this post I want to report on all of the other exciting science being done by my colleagues, after the great introduction by the programme director, Professor Stefan Marciniak. Elizabeth Le - Machine Learning to Predict Stroke Risk The very first talk of the day! Elizabeth gave a wonderful talk on her work in using supervised machine learning on CT scan data from carotid arteries, and how CNNs (convolutional neural networks) can be used to classify plaques that have a high probability of going on to cause disease vs benign plaques. Hopefully this can one day help us prevent strokes in at-risk patient groups! 🤖 Marciniak Lab💙 @MarciniakLabFirst @CambridgeMBPhD talk from @elostling“Radiomics & machine learning for the prediction of cardiovascular events”6 01:10 PM · Nov 21, 2019 Emma Rocheteau - Graph Neural Networks to Process Electronic Patient Records Electronic patient records are full of sparse data, often in an incomplete form, and so it was exciting to see Emma's work in using graph neural networks and graph transformer networks (a type of neural network) to process this kind of data, and use it to predict length of stay in hospitals! 📊 Marciniak Lab💙 @MarciniakLabEmma Rocheteau @CambridgeMBPhD describes her work using machine learning to examine patient electronic records1 04:57 PM · Nov 21, 2019 Minaam Abbas - A new immunotherapy for leukaemia My pal Minaam's talk focused on his work in investigating the mechanism of action of a "failed" drug abandoned by a large pharmaceutical company, a PROTAC for PCAF/GCN5. He was able to show some very interesting results on how in certain cell lines of acute myeloid leukaemia (AML), this drug actually causes the leukemia cell proliferation to slow by inducing their differentiation into dendritic cells, which have the potential to then present their own neo-antigens to the adaptive immune system, which has the potential to prevent relapses! 🧬 Souradip Mookerjee - Supercharged platelets I won't talk too much about my own work here, as I have written more about it on its own dedicated project page 😊 Marciniak Lab💙 @MarciniakLabSouradip Mookerjee @CambridgeMBPhD talking next about his work to engineer better blood platelets6 01:55 PM · Nov 21, 2019 Myrto Vlazdaki - Mathematical modelling of bacterial infection Myrto's talk started very philosophically discussing the role of models in medicine, and how she models the in vivo dynamics of Salmonella infection, which might otherwise be impossible to study by taking certain measurements that are more easily accessible and inferring states within the body that would not be easily directly measured! 🦠 Marciniak Lab💙 @MarciniakLabMyrto Vlazaki @CambridgeMBPhD - in her MB PhD she is using mathematical modelling to understand the in vivo dynamics of Salmonella infection6 02:20 PM · Nov 21, 2019 Joachim Hanna - Hedgehog signalling in T helper cell differentiation Joachim's talk was on the role of a type of intracellular signalling (the Hedgehog pathway) in the normal control of T cell differentiation. He was able to show that this type of signalling is key for the type of T cell (immune response) that protects us against fungi (Th17), that this goes wrong in inflammatory bowel diseases, and that modulating this pathway with drugs might help alleviate a lot of the pathology of these autoimmune conditions! 🦔 Marciniak Lab💙 @MarciniakLabFirst @CambridgeMBPhD talk from @elostling“Radiomics & machine learning for the prediction of cardiovascular events”6 01:10 PM · Nov 21, 2019 Maria Fala - 13C MRI for Metabolic Imaging of Brain Tumours Maria's talk focused on using this special type of MRI to measure the metabolic activity of glioblastomas (a type of brain tumour), something traditionally done with PET that involves radiation. This could pave the way to more non-invasive, safer ways of imaging tumours and monitoring how effective different treatments are, as well as for discovering metastases! 🔬 Summary Overall, this is a very exciting time for medical research, and many things that seemed like science-fiction only a few years ago now seem the new normal, and progress only seems to be getting more fast-paced! Everyone did a great job and I am proud to call them my friends 🙂 (and I haven't even mentioned the poster presenters and the audience for the insightful questions that were asked). 2019-08-03T00:00:00-00:00 https://souradip.com/blog/haemoglobin-assay (i.e. how to "pirate" your chemicals) I work in a Hematology lab. Often in research we use kits to save time, but they can often be eye-wateringly expensive and difficult to debug. Recently I had to measure the hemoglobin concentration in a large number of samples. These were very dilute, so sticking them in the normal blood analyzer machine did not work so well. I needed to do this manually. Here's a story about how I managed to make a hemoglobin assay kit (normally $300!) for basically nothing. Hopefully things like this will speed up research and help us discover better life-saving treatments! What I had available to me a plate reader (to take lots of measurements of absorbance quickly) and the usual reagents around any biology lab. Aims To measure hemoglobin in really diluted samples (not regular blood samples) To do this for lots of samples quickly and cheaply If I had lots of the reagent required, I wouldn't have to worry about running out or ordering more (that's more admin work than lab work and I'd much rather do the latter!) History of Hemoglobin measurement Hemoglobin was one of the first blood parameters to be measured routinely. There are lots of ways to do it! In this case, I needed something fast and scalable (on a 96-well plate reader). Normally, the absorbance spectrum of hemoglobin really varies, because it changes depending on if it bound to oxygen and various other substances. The goal of any reagent to help with this is to stabilise this to only one form, and hence one peak of hemoglobin in the absorbance spectrum. The classic is the cyanide-based Drabkin's solution that turns hemoglobin into a form that has a well-defined spectrum. However, this is not very stable over time, contains cyanide and not usually available in most labs. In recent times, there have been many more methods at measuring hemoglobin, from using Sodium laurel sulphate (AKA Sodium dodecyl sulphate, SDS, which is common in Western blotting as a detergent) to HARBOE. There's actually a pretty good review on the topic here, but I will focus on one method that I found to be quick, cheap and effective. The Kit I started with The QuantiChrom Hemoglobin Assay kit (catalogue no.: DIHB-250 from BioAssay Systems) was the recommended kit to do this with. The manufacturer has recently become eco-friendly, which basically means they don't print out the instructions for you and put them in the box. The kit actually comes with two small bottles. A 50ml bottle labelled "Reagent" and a 10ml bottle labelled "Calibrator". That's it. $309 please. Enough for 250 tests. This sounds like a lot, except everything needs to be done in duplicate on a plate reader, so it's really only 125 tests. Minus the calibrator and the water (not included) samples, it was only about enough for all of my samples (~100 of them) - from one experiment! The bottle labelled "reagent" ran out first, since I only needed 200uL of the calibrator once here, as the manufacturer has helpfully demonstrated how the kit is linear up to 200mg/dL! Initial Research I read carefully in the datasheet, and it says it is "based on an improved Triton/NaOH method". That sounds interesting. I dug up a few papers on the "normal" Triton/NaOH method. Triton is a detergent commonly used in Biology and NaOH is, well, sodium hydroxide. Presumably, the Triton is to lyse the red blood cells and the NaOH is the make the pH such that the hemoglobins oxidise and form a single peak in the spectrophotometer (rather than the multiple peaks that the different forms of hemoglobin when bound to oxygen give). I found a paper on a reagent called "AHD575" which is a Triton/NaOH mixture (it stands for "alkaline hematin + detergent") that should be measured at 575nm. I wondered if the "improvement" came from the instructions in the kit that said to measure at 400nm instead. It's possible to buy this for about €36, but we can do better by just using common lab reagents (already available almost everywhere) making this basically free. Testing this hypothesis So, I ran one of my samples both using the original reagent in the kit as well as the one I made up in the lab using 0.1mol/l NaOH and 2.5% Triton (to make up 100ml, we just need 0.4g NaOH [remembering from Chemistry that the mass to add = conc (M) x Mr x vol (litres)] and add 2.5ml Triton if in a liquid form or 2.5g if in a solid form). I also did a spectral scan to see if the reagent did anything different to the rest of the spectrum of the hemoglobin sample. See the results for yourself! Conclusions So, to make your own reagent for a hemoglobin quantification kit equivalent to DIHB-250/AHD575, you just need to add: 0.1mol/l NaOH 2.5% TritonX-100. e.g. to make 100ml of this, just add: 2.5ml TritonX-100 0.4g NaOH 97.5ml dH2O Instructions: (same as kit) Add 50uL of sample in duplicate to a 96 well plate Add 200uL of the reagent to each well Incubate 5mins at room temperature Read absorbance at 400nm on a plate reader The calibrator you can actually make up pretty easily too, if you have access to a blood testing machine - just test some blood on there, then dilute it as the kit suggests for using in the assay, and voila, a reference point! Since the absorbance varies linearly over the concentration range suggested in the kit, this should be enough to get a more standardised value of hemoglobin concentration! 2019-05-26T00:00:00-00:00 https://souradip.com/blog/hello I thought this would be a good place to put some my own personal thoughts, informal findings and other various musings. Hopefully I will be able to share the odd bits and pieces that I think might be useful to others that I can't put easily anywhere else! Stay tuned! 2020-06-17T00:00:00-00:00 https://souradip.com/projects/phd 📊 Poster · 📕 Thesis · 📄 Paper (in progress) Platelets are the smallest type of cells in our blood, and we have on average about 450 x 109/L of them. They are most well-known for their role in clotting, have no nucleus, but do contain mitochondria and lots of granules. When they are activated after contact with tissue damage, platelets release their granular contents. This mechanism has the potential to be used as a targeted drug delivery system. We used platelets made from stem cells in vitro that had their α-granules engineered, either genetically through lentiviral transduction or passively, through endocytosis, to contain Factor VIIa, a recombinant clotting factor and tested them in mouse models. This is normally commercially available as a drug for severe bleeding that usually also demonstrates significant thrombotic side effects. We aim to investigate whether platelets can mitigate these side-effects while still showing therapeutic efficacy. Summary Platelets can be made from stem cells genetically engineered with FVIIa into their granules Mouse models can be used to measure the effect of transfusion of human platelets 2019-10-01T00:00:00-00:00 https://souradip.com/projects/ed 📊 Poster · 📄 Paper (in progress) Emergency department (ED) crowding is a leading issue in many EDs and lead to worse patient outcomes. The ability to accurately forecast the occupancy and breach performance in ED a signicant time in advance would be operationally useful, allowing for better allocation of staffing resources, while accurate prediction of admissions would allow for more effective bed management. In a health system stretched for resources, accurate forecasts would allow us to more effectively use the resources we already have. In this study, we used the Addenbrooke's Hospital electronic health record system (EHR) to collect minute-by-minute data over a two year period, and used this train a deep neural network to accurately forecast ED occupancy, admissions and breach performance for each hour in the subsequent 24 hour period and evaluated its performance compared to existing techniques. Summary Created a dataset far more extensive and high-resolution than any previous dataset on hospital admissions (minute-by-minute for two years) Deep neural networks are a very good tool for predictive forecasting of ED demand and hospital admissions. Able to pull out non-linear patterns from the previous 96 hours and performs much better than traditional statistical techniques. 2018-12-01T00:00:00-00:00 https://souradip.com/projects/falls 🌍 Website · 📄 Paper (in progress) A medical IoT Device that uses on-board embedded deep learning (making this privacy-conscious, saving on internet usage and increasing the speed of response) to listen for the sound of elderly people falling over, and calls for help on their behalf. This solution is fast (real-time), supports swarming (multiple units in different rooms, e.g. in care homes), two-way communication, other forms of passive non-invasive monitoring and is cheap to manufacture. Currently being validated in care homes to more effectively use existing limited staffing resources to look after larger numbers of elderly people as well as being field-tested for direct-to-consumer sales. Summary Won a prize for best talk at the Royal Society of Medicine Patient Safety Conference. Part of my journey on the NHS Clinical Entrepeneur programme. 2016-08-01T00:00:00-00:00 https://souradip.com/projects/cenpj 📄 Paper (in progress) Primordial dwarfism is a rare form of dwarfism where babies are born small and stay small. One gene that causes this when mutated, both in mice and humans, is Cenpj, but the mechanisms how this occurs was unclear. By understanding what regulates growth normally, we might gain an insight into how this becomes dysregulated in cancer. The rest of this story is currently embargoed, check back later for more details!