covid-bioinformatics

Software tools to collect and analyze Coronavirus (COV) sequences, including code that extracts gene and protein sequences from COV genomes from GenBank, and creates gene and protein-specific sequence collections, alignments, and Hidden Markov Models (HMMs) for protein and nucleotide COV sequences. The code uses the cov_strains.yaml file that lists the expected genes for SARS, MERS, and COV2 so that collections are strain-specific.

The COV genes and proteins are parsed from the GenBank files as features and assigned standard names based on their product or mat_peptide tags. Possible synonyms for these standard names are listed in cov_features.yaml. A QC step compares the COV sequences to expected lengths listed in the cov_length_variants.yaml file.

example: create nucleotide and protein HMMs

• mkdir COV2
• download_gb_by_taxid.py
• feature_to_gene_and_protein.py COV2/*.gb
• seqs_to_aligns_and_hmms.py COV2/*.fa

software requirements

• Python3
• Biopython
• A sequence aligner, muscle, or clustalo, or mafft) (default: clustalo)
• HMMER 3.3

Code is included that annotates COV genomes using the nucleotide HMMs, pyTMHMM, and Infernal for RNA motif prediction and creates BED files for genome visualization.

example: annotate a COV genome by creating a BED file

• annotate_to_bed.py -rfam_file ../COV2/cov_allvirus.cm COV2/NC_045512.2.gb

optional software to make BED files

• Infernal
• RFAM COV for RNA motif prediction
• pyTMHMM for trans-membrane region prediction

recommended for best performance

• Get an NCBI API key
• Pass the key as a command-line argument or configure as an env variable (e.g. export NCBI_API_KEY=exampleexampleexampleexampleexample)

output directory

You can specify a destination directory with -data_dir, and the strain-specific directories (e.g. 'COV2') will be created inside this directory. The data_dir location is the current directory by default, and the default strain is COV2.

file formats created

• GenBank sequence: .gb
• Fasta sequence: .fa
• Fasta index file: .fai
• Stockholm alignment: .sto
• If the aligner is clustalo: Fasta alignment: .fasta
• If the aligner is mafft: MAF alignment: .maf
• HMMER profile HMM: .hmm
• hmmsearch table: .tblout
• BED track file: .bed