Tool for comparing two sets of gene families

This tool compares genefamily sets built from two different sets of HMM databases to give family correspondences between the two sets. This tool can be used to compare two family sets built at different levels in the species tree to detect large scale duplications or deletions. This can also be used to compare families obtained from different family building tools.

System requirments

• Unix based OS
• PYTHON 2.7
• hmmpress and hmmscan tools from HMMER package installed and added to PATH
• PYTHON modules re, sys, os, subprocess and argparse

Tool options

usage: compare_family_sets.py [-h] --hmm1 SET1_HMM --hmm2 SET2_HMM --fasta
                              FASTA_FILE [--e HMMSCAN_EVAL_CUTOFF]
                              --output_dir OUTPUT_DIR --n1 SET1_NAME --n2
                              SET2_NAME [--t NUM_THREADS]

Tool for comparing two family sets through family HMM sets

Arguments:
  -h, --help            show this help message and exit
  --hmm1 SET1_HMM       Concatenated HMM file for first set of families
  --hmm2 SET2_HMM       Concatenated HMM file for second set of families
  --fasta FASTA_FILE    Fasta file containing sequences that will be clustered
                        into both set of families
  --e HMMSCAN_EVAL_CUTOFF
                        hmmscan E-value cutoff
  --output_dir OUTPUT_DIR
                        Location of output directory
  --n1 SET1_NAME        Name for the first set of families
  --n2 SET2_NAME        Name for the second set of families
  --t NUM_THREADS       Number of threads for hmmscan

Output files

• The final output is written in files <n1>-<n2> and <n2>-<n1> that contain the correspondences for set1 vs set2 and set2 vs set1 respectively.
• Format: <first-set-famid> <second-set-famid> <number-sequences-common-between-two-families>

Intermediate output files

• <n1>.famlist and <n2>.famlist contain the respective family sets. Format: famid sequenceid
• <n1>.hmmtblout and <n2>.hmmtblout contain the respective hmmscan results
• <n1>.hmm.* and <n1>.hmm.* are the respective hmm database index files