Table of Contents

• Overview
• Installation
• Input files
• Walk-through
  • Input Sanitation
  • RPG digest
  • Peptide Processing
  • Coding Sequence Extraction
  • General Summary Statistics
  • Genomic co-ordinate conversion
  • Exon-exon Junction Covering Peptide Identification & Summary Statistics
  • Unique Peptide Counter
• Contact

Overview

ProtView is designed to present statistics of in silico digestions and provide useful information, such as the protein sequence coverage, peptides covering exon-exon junctions, and the percentage of junctions or residues in the data that are covered by peptides of a digest (Figure 1). It allows the users to see the portions of the gene/transcript on the genome that are covered by proteomic data. It takes the protein sequence (fasta) and the coding sequence annotations (gff3 format) on the genome as inputs. It utilises Rapid Peptides Generator (RPG) (Maillet, 2019), which carries out the in-silico digestion. It then maps the digested proteins back to transcripts/genes on the genome, which allows the comparisons of the transcript/ gene sequences visible to proteomics experiments between different digestions.

Figure 1: outline of the ProtView workflow

Installation

Supported Python version: >= 3.8

ProtView requires the following packages:

• rpg
• pandas
• gffpandas
• numpy
• BioPython

We recommend creating a separate environment to run ProtView in the command line, to avoid clashes between package versions in the main environment. To create and activate a separate environment named 'protview_environment', install ProtView and dependencies, and run setup commands:

conda create -n protview_environment
conda activate protview_environment
pip install protview
protview setup_commands

Setup commands include the creation of two user-defined proteases in RPG, 'Asp-N-UD' (cleaves N-terminal to D), and 'Glu-C-UD' (cleaves C-terminal to E). It was found in ProtView benchmarking analyses and comparisons to public data that these cleavage rules better represent the behaviour of these proteases.

Input Files

Required input for an analysis with ProtView:

• Protein sequence(s) in fasta format
• Corresponding gff3 file(s)

Any additional input required for downstream functions is generated by ProtView during the analysis.

Walk-through

ProtView is intended to be used in the command line. To view the available ProtView modules use protview -h. To get help for an individual module use protview <module name> -h

The Arabidopsis thaliana genes AT1G666600 and AT1G66610 (Figure 2) are included as example data to walk through how ProtView works. Araport11 protein sequences were retrieved from the tair database in fasta format and the Araport11 GFF3 files were downloaded via jbrowse. These files can be found and downloaded from the protview_example_data folder on the github page. Once the working directory has been set to the directory containing the intended input files, example data or otherwise, the analysis can begin.

Figure 2: Jbrowse depiction of the example

Input Sanitation

This step replaces underscores and '>' characters in the fasta protein descriptions to avoid downstream errors in the analysis. To run the fasta input sanitation function of the fasta file in the example data:

protview fasta_input_sanitation at1g66600_at1g66610.fasta

RPG digest

ProtView utilises Rapid Peptides Generator (RPG) (Maillet, 2019) to carry out in-silico digests. RPG takes a fasta protein sequence as input, more instructions for running RPG can be found here.

RPG can carry out digests in sequential (default) or concurrent mode (-d c), where a sequence is cleaved by multiple enzymes. ProtView can then combine peptides from sequential digests, creating parallel enzyme combinations.

To digest the example proteins with Trypsin, Asp-N, and Glu-C and save the output as 'at1g6600_at1g66610_rpg.fasta': rpg -i at1g66600_at1g66610.fasta -o at1g66600_at1g66610_rpg.fasta -e 2 23 42 -q. The argument -d c can be added if carrying out a concurrent digest with the enzymes, rather than sequential, and concurrent digests are then treated the same as single enzyme digests throughout the workflow.

Peptide Processing

This module processes RPG-generated peptides in fasta format, allowing the user to 1) specify the number of missed cleavages allowed per peptide, 2) create parallel enzyme digests, 3)filter for peptides containing a specific residue, 4) filter by a certain amino acid length.

RPG treats mis-cleavage as a percentage of how frequently cleavage is missed at each theoretical cleavage site. The aim of ProtView is to give the theoretical upper limit of peptides that could be identified in an experiment and allow for in silico results to be compared to experimental data from software algorithms that treat mis-cleavage as the number of missed cleavages allowed per peptide . A function for generating mis-cleaved peptides is therefore included in ProtView, which allows the generation of mis-cleaved peptides by concatenating adjacent peptide sequences from the RPG digest up to n times, where n is the mis-cleavage number specified by the user.

Required Arguments:

• input_file: name of the RPG peptides file

Options:

• -mc, --miscleavage: Mis-cleavage value, default = 0
• -e, --enzymes: Enzymes to create a parallel digest with, default = none
• -r, --residue: Residue to be filtered for, default = none
• -min, --min_len: Minimum peptide length to filter for, default = 7
• -max, --max_len: Maximum peptide length to filter for, default = 35

The output is saved after each of the above steps.

Output files are named <input_file>_<parameter_suffixes>.csv (e.g. at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel_len_7_35.csv) Parameter suffixes:

• mc<x>, where x = the number of missed cleavages, if mis-cleavage is used
• <enzymes>_parallel, if a parallel digest is created
• len_<min length>_<max length>, if filtered for amino acid length
• <residue> , if filtered for a specific residue

Output columns:

• FASTA_description: Description of the protein from th original FASTA file
• cleavage_position: Protein end coordinate of the peptide
• peptide_size: peptide length in amino acids
• mol_weight: molecular weight
• isoelectric point
• sequence
• peptide_start: Protein start coordinate of the peptide

To filter the individual enzyme digests, with default parameters. protview rpg_output_processing at1g66600_at1g66610_rpg.fasta

To create a parallel digest of Trypsin and Asp-N using default parameters: protview rpg_output_processing at1g66600_at1g66610_rpg.fasta -e Trypsin Asp-N

Notes:

• When creating a parallel digest, the enzymes need to entered exactly as they are in the RPG output (e.g. 'Asp-N' and not 'aspn', asp-n', 'Asp-n' etc.)
• Any files containing peptides from parallel digests or containing missed cleavages will have 'parallel' or 'mc' in their file names. It is important not to rename these files, as the statistics are calculated differently for mis-cleaved peptides and combined digests.

Coding Sequence Extraction

Extracts coding sequence (CDS) information from gff3 files and prepares this information for the downstream analysis. The resulting tables retain relevant information from the original gff3 file, with the addition of relative protein sequence coordinates of the CDS start and end positions, unique CDS IDs, and the lengths and unique IDs of introns in between CDSs. Relative protein sequence coordinates are calculated and CDS and intron IDs consist of chromosome, start position, end position, and strand.

Required Arguments:

• input_file: input file name of gff3 file

To extract CDS information from the gff3 file for the example Arabidopsis proteins:

protview cds_extraction at1g66600_at1g66610.gff3

The resulting CDS information is saved separately for each strand in csv format ('at1g66600_at1g66610_+_cdsdf.csv, at1g66600_at1g66610_-_cdsdf.csv,')

Output columns: Information is retained from the original gff3 files under the same columns names. Columns added by ProtView are:

• cds_id: unique ID of the CDS
• intron_id: unique ID of the adjacent intron before the CDS
• intron_length: length of the adjacent intron
• cum_intron: cumulative length of introns before the CDS
• cds_first_start: start position of the first CDS of an isoform
• protein_start: protein start coordinate of the CDS
• protein_end: protein end coordinate of the CDS

Note:

• ProtView will generate a CDS information file for each DNA strand, regardless of whether the input only contains protein(s) on one of the strands. It is important to keep empty CDS information files generated, as they are required as input for steps in the downstream analysis.

General Summary Statistics

This gives a table with columns for the total number of peptides generated, both before and after filtering, peptide length distributions, and protein sequence coverage. The calculation of residue coverage is optional and carried out if residues are provided by the user.

Required Arguments:

• -fasta, --fasta_files: original fasta sequence files used in the digests
• -u, --unfiltered_rpg_files: unfiltered rpg peptides in csv format (output from rpg_output_processing module)
• -f, --filtered_rpg_files: filtered rpg peptides in csv format (output from rpg_output_processing module)

Optional Arguments:

• -o, --output_name: name of the output file, ending in .csv. Default = summary_table.csv
• -r, --residues: residues to calculate coverage for

Output columns:

• enzyme: the protease or protease combination used in the digest
• total_peptides: the total number of peptides generated by that enzyme
• mean_length: mean length of all peptides generated by this enzyme
• median_length: median length of all peptides generated by this enzyme
• filtered_peptides: number of peptides that meet filtering criteria
• coverage: coverage % of the original digested protein sequence

To generate summary statistics of the example data for the individual and parallel digests, including Cysteine (C), Serine (S) and Threonine (T) coverage:

protview summary_stats -fasta at1g66600_at1g66610.fasta -u at1g66600_at1g66610_rpg.csv 
at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel.csv -f at1g66600_at1g66610_rpg_len_7_35.csv 
at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel_len_7_35.csv -r C S T

Note:

• Please ensure that any files containing peptides with missed cleavages or parallel digests have 'mc' or 'parallel' in their file names (standard ProtView output unless they have been manually changed by the user) before carrying out this step

Genomic co-ordinate conversion

To map peptides onto the genome, relative peptide coordinates from the digest output are converted to the outer bounds of their corresponding coordinates on the genome. The resulting table contains the isoform, both genomic and relative protein start and end positions for each peptide, and the enzymes used to generate the peptide.

This function works best on one protein at a time and the output can be visualized on the genome using tools such as Gviz.

Required Arguments:

• rpg_file: csv file containing the peptides and their proteomic co-orindates (output from rpg_output_processing module)
• cds_files: both csv files containing extracted coding sequences for each DNA strand, ending in +/-_cdsdf.csv (output from cds_extraction module)

To calculate, the genomic coordinates of the filtered parallel Trypsin:Asp-N digest in our example:

protview gen_coords at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel_len_7_35.csv 
at1g66600_at1g66610_+_cdsdf.csv at1g66600_at1g66610_-_cdsdf.csv

Output columns:

• isoform: the isoform that the peptide originates from
• Protein start coordinate: relative start coordinate on the protein sequence
• Genomic start coordinate: calculated start coordinate on the genome
• Protein end coordinate: relative end coordinate on the protein sequence
• Genomic end coordinate: calculated end coordinate on the genome
• enzyme: the protease or protease combination used in the digest

Note:

• The purpose of this function is to enable peptide visualisation on the genome and it works best when ran on small sets of input

Junction-covering peptides & statistics

Peptides are filtered for those that cover exon-exon junctions. The recommended input for identifying junction-covering peptides is the filtered digest results. Positive outcomes are saved in the same csv format as the digest results, with an additional column for junction location.

Required Arguments:

• rpg_file: csv file containing the peptides to be filtered (output from rg_output_processing module)
• cds_files: both csv files containing extracted coding sequences for each DNA strand, ending in +/-_cdsdf.csv (output from cds_extraction module)
• output_name: desired name of output csv file, ProtView will automatically differentiate between DNA strands

To filter the individual digests from the example

protview junction_peptides at1g66600_at1g66610_rpg_len_7_35.csv at1g66600_at1g66610_+_cdsdf.csv 
at1g66600_at1g66610_-_cdsdf.csv single_digest_junction_peptides.csv

To identify splice junction covering peptides from the parallel digest

protview junction_peptides at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel_len_7_35.csv 
at1g66600_at1g66610_+_cdsdf.csv at1g66600_at1g66610_-_cdsdf.csv parallel_digest_junction_peptides.csv

By giving 'single_digest_junction_peptides.csv' as the output argument, ProtView adds '_+_' or '_-_' to the end to distinguish between strands, resulting in files named 'single_digest_junction_peptides_+_.csv' and 'single_digest_junction_peptides_-_.csv'

Output columns from this step are the same as the peptide processing step, with the addition of:

• junction: the protein coordinate of the junction that is covered

A summary table can be generated from the junction-covering peptide results. Each table includes the number of junction-covering peptides generated by each enzyme, the number of unique junctions that an enzyme covers (to avoid double counting of splice junctions shared between transcripts), and a junction coverage percentage, which is the percentage of the total junctions available in the isoforms being examined that are covered by an enzyme.

Required Arguments:

• -pept, --junction_spanning_peptides: csv files containing the junction covering peptides
• -cds, --cds_files: both csv files containing extracted coding sequences for each DNA strand, ending in +/-_cdsdf.csv Optional Arguments
• -out, --output_name: desired name of output csv summary file, default: junction_summary.csv

To summarise the junction-covering peptides identified above:

protview junction_summary -pept single_digest_junction_peptides_+_.csv single_digest_junction_peptides_-_.csv 
parallel_digest_junction_peptides_+_.csv parallel_digest_junction_peptides_-_.csv -cds at1g66600_at1g66610_+_cdsdf.csv 
at1g66600_at1g66610_-_cdsdf.csv

Output columns:

• enzyme: the protease or protease combination used in the digest
• junction_spanning_peptides: total number of peptides that cover splice junctions
• unique_junctions_covered: number of unique junctions covered (to avoid double counting of those shared between isoforms)
• total_junctions_covered: the total number of junctions covered by peptides
• total_junction_coverage: % of available junctions that are covered by peptides

Unique Peptide Counter

ProtView can count the number of unique peptides per digest and append the calculation as a column to either the general proteomic summary table or the junction-covering peptide summary table. Unique peptides are defined as those with sequences that can only be found in one isoform in a digest.

Required Arguments:

• table_name: the proteomic or junction summary statistics table for the column to be appended to
• rpg_files: the filtered digest results in csv format (output from rpg_output_processing module)

To calculate the number of unique peptides in the example digest and append it to the proteomic summary table:

protview unique_count summary_table.csv at1g66600_at1g66610_rpg_len_7_35.csv 
at1g66600_at1g66610_rpg_Trypsin_Asp-N_parallel_len_7_35.csv

Output column: This column can be appended to either summary table generated

• isoform unique peptides

Contact

For further information please contact SSPuliasi@dundee.ac.uk