Access Infernal commands from Julia.

This package is registered. Install with:

import Pkg
Pkg.add("Infernal")

This package does not export any symbols.

Use either qualified calls such as Infernal.cmfetch(...) or import the functions you need explicitly:

using Infernal: cmfetch, cmalign, cmalign_parse_sfile

Most functions are thin wrappers around Infernal command-line tools. They write command outputs to temporary files and return a named tuple containing the generated paths.

• cmfetch: fetch a covariance model from a CM database
• cmalign: align sequences to a covariance model
• cmalign_parse_sfile: parse cmalign --sfile output into a DataFrame
• cmbuild: build a covariance model from an alignment
• cmemit: emit example sequences from a covariance model
• cmcalibrate: calibrate a covariance model
• cmsearch: search sequences with a covariance model
• cmsearch_parse_tblout: parse cmsearch --tblout output into a DataFrame
• esl_afetch: fetch an alignment from an MSA database
• esl_reformat: convert between sequence/alignment formats

using Infernal: cmfetch, cmalign, cmalign_parse_sfile

cm = cmfetch("/path/to/Rfam.cm", "RF00162")
alignment = cmalign(cm.out, "/path/to/sequences.fasta"; outformat="Stockholm")
summary = cmalign_parse_sfile(alignment.sfile)

summary[:, ["seq_name", "bit_sc", "avg_pp"]]

alignment.out contains the alignment written by cmalign, while alignment.sfile contains the per-sequence summary parsed above.

using Infernal: esl_afetch, esl_reformat

seed = esl_afetch("/path/to/Rfam.seed", "RF00162")
afa = esl_reformat("AFA", seed.out; informat="Stockholm")

println(read(afa.out, String))

using Infernal: cmbuild, cmemit, cmsearch, cmsearch_parse_tblout

built = cmbuild("/path/to/alignment.sto"; informat="Stockholm")
emitted = cmemit(built.cmout; N=5, outformat="AFA")
search = cmsearch(built.cmout, emitted.out)
hits = cmsearch_parse_tblout(search.tblout)

• The wrapped Infernal commands must succeed or the functions throw ProcessFailedException.
• Temporary output files are not deleted automatically; remove them when they are no longer needed.

Supported platforms:

• Linux
• macOS (Intel & M-series)

Windows not supported because Infernal is not available for Windows (use WSL instead).