pepti_map

pepti_map is a tool for mapping peptide sequences to their possible genomic loci. It does so based only on the sequence information, so that no further peptide information besides the amino acid sequence and no annotation of the genome is required, enabling a mapping to personal genomes. pepti_map utilizes RNA-seq reads of the same sample to facilitate the mapping: Each peptide is first matched to RNA-seq reads, which are then assembled into longer contigs and aligned onto the genome.

Setup

To setup pepti_map, first install all dependencies listed in the environment_<os>.yml (environment_macos.yml or environment_linux.yml). We recommend using Mamba. For example:

$ mamba env create -f environment_linux.yml

pepti_map relies on the following tools to be installed:

• Trinity
• GMAP
• PoGo

If you install the dependencies via the given environment_<os>.yml, both Trinity and GMAP should already be installed. To install PoGo, download the latest release from its GitHub page.

You will then need to add the path to the directory in which your PoGo installation is located in an .env file via the POGO_PATH environment variable, e.g.:

POGO_PATH=/Users/me/Tools/PoGo_v1.2.3/Linux

This is the only environment variable that needs to be set for pepti_map to work. You can, however, set additional environment variables. Below you will find a table listing all environment variables.

Environment Variable Name	Usage
POGO_PATH	The path to the directory in which the PoGo installation is located.
IO_N_PROCESSES	The number of processes to use when generating the input files for PoGo. If not set, defaults to multiprocessing.cpu_count().
TRINITY_USE_DOCKER	Whether to run a dockerized version of Trinity. Value must be True or False. If not set, defaults to False. If set to True, a dockerized version of Trinity must be installed on the system.
TRINITY_PATH	The path to the Trinity installation. If not given, expects Trinity to be executable from the working directory (e.g. by using an installation via a Mamba environment).
TRINITY_N_PROCESSES	The number of processes with which to run Trinity in parallel. If not set, defaults to multiprocessing.cpu_count() // TRINITY_N_CPUS (floor division).
TRINITY_N_CPUS	The number of CPUS to use for one Trinity process. Corresponds to the --CPU option of Trinity. If not set, defaults to 2.
TRINITY_MAX_MEM	The max memory to use for one Trinity process. Corresponds to the --max_memory option of Trinity. If not set, defaults to "1G".
GMAP_N_THREADS	The number of threads with which to run GMAP during the alignment. Corresponds to the -t option of gmap. If not set, defaults to multiprocessing.cpu_count().
GMAP_BATCH_MODE	The batch mode in which to run GMAP during the alignment. Corresponds to the -B option of gmap. If not set, defaults to 2.
TEMP_DIR_PATH	The path to the folder in which temporary results are saved. If not set, defaults to ./temp.

Usage

To run pepti_map, first activate your Mamba environment.

The general usage of pepti_map is as follows:

$ (peptimap-env) python -m pepti_map.main [OPTIONS]

A minimal command would look like this:

$ (peptimap-env) python -m pepti_map.main -p path/to/peptide/file -r path/to/rna/reads/file -g path/to/genome/fasta

In order for pepti_map to be able to run, at least the -p and -r options need to be set, as well as one of the -g or -x option.

For further specification of these and further options, see the table below or use

$ (peptimap-env) python -m pepti_map.main --help

Overview of the options for running pepti_map:

Option	Usage
-p / --peptide-file	The path to the peptide file (for format, see below).
-r / --rna-file	The path to the RNA-seq file. In case of paired-end sequencing, this file is expected to be in forward orientation.
-pa / --paired-end-file	The path to the second RNA-seq file in case of paired-end sequencing. This file is expected to be in reverse orientation. If none is given, the RNA-seq file given with the -r option is assumed to result from single-end sequencing.
-c / --cutoff	The position of the last base in the reads after which a cutoff should be performed (starting at 1). The cutoff is applied to all reads. If the value is equal to or smaller than 0, no cutoff is performed. To define different cutoff values for the RNA-seq files in case of paired-end sequencing, you can supply two cutoff values by using the -m option twice (e.g. -m 80 -m 60). The first value is used for the file supplied with -r, whereas the second value is used for the file supplied with -pa. (Default: -1)
-k / --kmer-length	The k-mer size used during the mapping of peptides to RNA-seq reads. As the RNA-seq reads are 3-frame translated for the mapping, the k-mer size refers to amino acids. (Default: 7)
-o / --output-dir	The path to the output directory for all generated files. (Default: ./)
-pi / --precompute-intersections	If used, the intersection sizes for the Jaccard Index calculation are precomputed during the matching phase.
-j / --jaccard-index-threshold	Sets of matched RNA-seq reads per peptide will only be merged together if their Jaccard Index has a value above the given threshold. (Default: 0.5)
-m / --merging-method	Which merging method to use for sets of matched RNA-seq reads. Must be one of agglomerative-clustering, full-matrix. (Default: full-matrix)
-cl / --min-contig-length	Sets the --min_contig_length option for Trinity during assembly. A value below 100 is not possible. (Default: 100)
-g / --genome	The path to the genome file(s) to align to. In case of multiple files, the paths must be separated by comma.
-x / --gmap-index	The path to an existing GMAP index that should be used instead of building a new one. If this option is set, the -g / --genome option is ignored.

Input file formats

Peptide file ( -p / --peptide-file):

This file should contain a list of peptides in form of amino acid sequences, with one peptide sequence per line. Optionally, the file may contain protein group information per peptide, with this information being on the same line as the peptide, separated by tab. A header is not needed. One line may thus look as follows:

<peptide sequence>

or

<peptide sequence>  <protein group information>

If the protein group information is given, peptides with the same protein group will be grouped together, with matches to the RNA-seq reads being allocated per group. If not given, each peptide is treated as a separate group.

RNA-seq file(s) (-r / --rna-file and -pa / --paired-end-file):

These files should contain the RNA-seq reads of the same sample/subject as the peptide data in FASTQ or gzipped FASTQ format.

Genome file(s) (-g / --genome):

These files should contain the genomic sequences to align to in FASTA format.

pepti_map output

pepti_map will output a GTF (pepti_map_output.gtf) and a BED (pepti_map_output.bed) file containing the mappings of the peptides to genomic loci. Peptides that were not matched will not appear in these files. For further information on these formats, please see the PoGo documentation, as these outputs correspond directly to the respective PoGo output formats.

Additionally, pepti_map outputs a quantification file (peptide_read_quant.tsv), containing the group id assigned to each peptide, the number of reads that were matched to the peptide, and the number of reads that were matched to the group the peptide belongs to. This means, for all peptides belonging to the same group, their individual read counts add up exactly to the read count for the group. In case a peptide is too short (smaller than the k-mer length), it will be assigned a group id of -1 and excluded from further calculations. The format per line is as follows, with the individual values being tab-separated:

<peptide sequence>  <numeric group id>  <number of matched reads for peptide>   <number of matched reads per group>