TAPscan (perl) script to detect TAPs (Transcription Associated Proteins) in your sequence data.

1. Install HMMER according to its installation instructions for your system.
2. Perl is required (script was tested with v5.34.0)

The test-data folder contains example data (input and output). To run TAPscan on this folder, use the tapscan_run.sh script.

./tapscan_run.sh test-data

Step 1: Filter dataset for isoforms if necessary

This script can be used if you have a fasta file containing isoforms. The first part of the IDs has to be the same up to a certain pattern which has to be given as a parameter. The script then groups all proteins with the same ID part together and uses the longest one as major isoform. All the other sequences will be written into an iso.fa file. The script also adds the letter code at the beginning of the fasta ID. The letter code has to be given using the -c option, the output directory using the -o option, the input fasta file using the -i option and the pattern used for splitting using the -p option.

python split_isoform_dot.py -o outputfile -I CAMSA.fa -c CAMSA -p .

Step 2: Run TAPscan Classify

TAPscan is a comprehensive tool for annotating TAPs with a special focus on species belonging to the Archaeplastida. In general, the detection of TAPs is based on the detection of highly conserved protein domains.

During the first step, each sequence out of a species protein set is scanned for protein domains (stored as profile Hidden Markov Models) using hmmsearch. The domains file (domains_v12.txt) consists of 154 profile HMMs and functions as the domain reference during the hmmsearch command. v Afterwards, by applying the TAPscan script (tapscan_classify.pl), specialized rules are applied to finally assign the protein sequences to TAP families based on the detected domains in the previous step. With the latest TAPscan v4, a protein set can be scanned for 137 different TAP families with high accuracy through applying GA-thresholds and coverage values

hmmsearch --domtblout sample.domtblout --cut_ga domains_v13.txt sample.fa

tapscan_classify_v4.76.pl \
sample.domtblout \
rules_v82.txt \
sample.output.1.family_classifications.txt \
sample.output.2.family_statistics.txt \
sample.output.3.subfamily_classifications.txt \
coverage_values_v11.txt

A script to run these commands on all FASTA files in a folder can be found in tapscan_run.sh

TAPscan provides the user with three different output file formats. Each output file is semicolon-separated. Output 1 contains the detected domains and finally assigned TAP family for each gene ID. If domains are assigned to a sequence but not all rules are fulfilled, the sequence is assigned to 0_no_family_found. Output 2 is a summary of the number of members for each TAP family. Output 3 is similar to Output 1 but contains additional information about subfamilies.

Output.1: Family Classifications

GENE ID;TAP_FAMILY;NUMBER_OF_FAMILIES_FOUND;DOMAINS;…
ARATH_AT1G01010.1;NAC;1;NAM;
ARATH_AT1G01030.1;ABI3/VP1;1;B3;
ARATH_AT1G01040.2;Dicer;1;PAZ;Ribonuclease_3;Helicase_C;dsrm;DEAD;
ARATH_AT1G01060.1;MYB-related;1;Myb_DNA-binding;
ARATH_AT1G01110.2;0_no_family_found;0;IQ;

Output.2: Family Statistics

TAP_FAMILY;NUMBER_MEMBERS
0_no_family_found;676
ABI3/VP1;64
ADA2;2
Alfin-like;8

Output.3: Subfamily Classifications

GENE_ID;TAP_FAMILY;SUBFAMILY;NUMBER_OF_FAMILIES_FOUND;DOMAINS;…
ARATH_AT1G01010.1;NAC;-;1;NAM;
ARATH_AT1G01030.1;ABI3/VP1;-;1;B3;
ARATH_AT1G01040.2;Dicer;-;1;PAZ;Ribonuclease_3;Helicase_C;dsrm;DEAD;
ARATH_AT1G01060.1;MYB;MYB-related;1;Myb_DNA-binding;
ARATH_AT1G01110.2;0_no_family_found;-;0;IQ;
ARATH_AT1G01160.2;GIF;-;1;SSXT