peptide-matcher

peptide-matcher is a piece of software that can be used for matching peptide sequences identified in proteomics experiments using a database-match or a de novo approach against a sequence database. The main purpose is to extract sequence context for the corresponding matches, but peptide-matcher can also provide structural context if provided with a database that includes structural information, see peptide-matcher-data.

There are three ways of how to use peptide-matcher:

• the GUI peptide_matcher_gui
• the CLI peptide_matcher
• the python class peptide_matcher.PeptideMatcher

The GUI is written with wxWidgets. Other dependencies include biopython and pyahocorasick.

Installation

Install with pipy: pip install peptide_matcher.

How to use the GUI

Two files are needed: the database in fasta format with optional structural annotations and a plain list of peptide sequences.

The optional structural annotations should follow a custom format. Databases generated based on alphafold's models for a couple of popular model organisms are distributed at peptide-matcher-data.

The results of the peptide matching are returned to the GUI and can be saved as xlsx. For each peptide the following output is generated:

Field	Description	Example	Values
Peptide	peptide sequence	QVHAVSFYSK	string of amino acid symbols
Length	peptide length	10	integer
Protein	matching protein id	A6NL46	string
Start	start position (1-based)	150	integer
End	end position (1-based)	159	integer
C-term	distance to protein's C-terminus	182	integer
N-flank	N-flanking residues in this protein	TDKA	string
C-flank	C-flanking residues in this protein	GHGV	string
N-flank*	weblogo for each position of the N-flank	2T|2D|2K|2A	| - separator between positions
C-flank*	weblogo for each position of the C-flank	1G1D|2H|1G1E|2V
N-flank SS	secondary structure for the N-flank	HHHH	string of DSSP codes
Peptide SS	same for the peptide itself	HH------EE
C-flank SS	same for the C-flank region	EEEE
N-flank TM	transmembrane region for the N-flank	TTTT	string of: T - TM region, S - signal peptide
Peptide TM	same for the peptide itself	TT--------
C-flank TM	same for the C-flank region	----
N-flank conf	alphafold's pLDDT score for the N-flank	43,46,40,49	list of integers 0-100
Peptide conf	same for the peptide itself	44,44,45,44,50,39,48,39,56,46
C-flank conf	same for the C-flank	49,47,42,46
N-flank RSA	relative solvent accessibility for the N-flank	81,79,84,71	list of integers 0-100
Peptide RSA	same for the peptide itself	90,78,75,78,54,62,73,84,73,81
C-flank RSA	same for the C-flank	67,78,71,80

In the provided database, the RSA values are calculated by dividing the absolute solvent accessibility (ASA) as produced by dssp (mkdssp v.3.0.0) by the theoretical maximum values for ASA from Tien et al 2013.

How to use CLI

Check out peptide_matcher -h:

$ peptide_matcher -h
usage: peptide_matcher [-h] --peptides FILENAME --database FILENAME [--secstruct] [--flanks N] [--format {json,tsv,csv}] [--output OUTPUT]

Match peptides in a protein database.

optional arguments:
  -h, --help            show this help message and exit
  --peptides FILENAME, -p FILENAME
                        list of peptides to match
  --database FILENAME, -d FILENAME
                        protein database in fasta format
  --secstruct, -s       whether the database also contains structural information
  --flanks N, -f N      length of the flanks to report (default: 4)
  --format {json,tsv,csv}, -F {json,tsv,csv}
                        output format (default: json)
  --output OUTPUT, -o OUTPUT
                        output file (default: output to stdout)

The output is similar to that of the GUI. The header of the tabular output formats looks as follows: [ 'peptide', 'peplen', 'record_id', 'start', 'end', 'c_term', 'n_flank', 'c_flank', 'n_logos', 'c_logos', 'sst_n_term', 'sst_pept', 'sst_c_term', 'tm_n_term', 'tm_pept', 'tm_c_term', 'conf_n_term', 'conf_pept', 'conf_c_term', 'acc_n_term', 'acc_pept', 'acc_c_term' ]. The json output is a list of dictionaries with each one of the following format: {"peptide": "IYGALAVGAP", "matches": [{"record_id": "P77549", "start": 157, "end": 166, "c_term": 227, "n_flank": "NGMA", "c_flank": "LGLL", "sst_n_term": "HHHH", "sst_pept": "HHHHHHHHHH", "sst_c_term": "HHHH", "tm_n_term": "----", "tm_pept": "----------", "tm_c_term": "----", "conf_n_term": [94, 89, 91, 94], "conf_pept": [93, 86, 88, 94, 89, 85, 90, 92, 86, 88], "conf_c_term": [93, 94, 91, 94], "acc_n_term": [3, 6, 25, 6], "acc_pept": [9, 24, 19, 0, 25, 50, 44, 0, 22, 45], "acc_c_term": [36, 0, 37, 47]}], "n_logos": [{"N": 1}, {"G": 1}, {"M": 1}, {"A": 1}], "c_logos": [{"L": 1}, {"G": 1}, {"L": 1}, {"L": 1}]}.

How to use the API

from peptide_matcher import PeptideMatcher, wrap_logos, wrap_scores
peptides = [ 'IYGALAVGAP', 'LTCDETPVFSGSVLN', 'KRFARESGMTLL', 'GAGFAELLSSLQTPEIK', 'RTGHKLV' ] # or a file handle
database = 'UP000000625_83333_ECOLI.fasta' # or a file handle
flanks = 4
secstruct = True
pm = PeptideMatcher(peptides, database, secstruct, flanks)
for output in pm.run():
    print(output)