NeoFlow2: a proteogenomics pipeline for neoantigen discovery

NeoFlow2 is a streamlined computational workflow that integrates WES and MS/MS proteomics data for neoantigen prioritization to facilitate cancer immunotherapy. It includes four modules:

• Variant annotation and customized database construction
• Variant peptide identification including MS/MS searching, FDR estimation, PepQuery validation
• Human leukocyte antigen (HLA) typing
• MHC-binding prediction and neoantigen prioritization.

The four modules are streamlined using Nextflow and Docker. NeoFlow2 supports both label free and iTRAQ/TMT data. NeoFlow2 could be run in a single Linux computer or Could environment like AWS.

How to run

• run with docker locally

nextflow run bzhanglab/neoflow2 -r main -profile docker \
   -params-file /path/to/my/parameter_file.yml

• run with aws batch on AWS cloud. For more information, please refer to Running nextflow on AWS cloud.

nextflow run bzhanglab/neoflow2 -r main -profile awsbatch \
   -bucket-dir s3://mybucket/workdir/2022-05-31 \
   -params-file /path/to/my/parameter_file.yml

To run with awsbatch, you must specify an s3 path as outdir, e.g. --outdir s3://mybucket/myfolder. In this case, results will be stored a directory under s3://mybucket/myfolder. The directory name is defined by the parameter run_version.

The parameter file is a yaml file that contains all the parameters. Below is a sample parameter file:

---
run_version: "2022_05_31_run"
outdir: "s3://mybucket/myfoler"
manifest: "/path/to/manifest/file.tsv"
maf: "/path/to/maf/file"
fusion_file: "/path/to/tared/gzipped/fusion/file"
bam_source: "uuid"
bam_type: "bam"

The following table lists the parameters and their default values if available.

parameter name	note
run_version	name of the current run
outdir	path to a local directory (/data/neoflow2_output) or s3 path (e.g. s3://mybucket/neoflow2_outpt)
manifest	path to input manifest file (see below for more detail)
maf	path to maf file which contains somatic mutations for all the samples included in the manifest file
fusion_file	path to tared and gzipped fusion file
bam_source	"uuid" (gdc) or "url" (http) or "path" (s3, or local file path), Default is "uuid"
bam_type	"bam" or "cram", Default is "bam")
database	path to database output of previously run for global proteomics (this is used for running phosphoproteomics only)
hlatyping	path to hla typing output of previously run for global proteomics (this is used for running phosphoproteomics only)
annovar_protocol	The parameter of "protocol" for ANNOVAR, default is "refGene". Find more about the setting of this parameter at https://annovar.openbioinformatics.org/en/latest/user-guide/startup/
annovar_anno_file	ANNOVAR annotation datafile. All required annotation files are needed to be in a single TGZ format file
annovar_file	ANNOVAR package file path. This is a TGZ format file which contains the ANNOVAR package.
annovar_buildver	The genome build version. For example, hg19 or hg38. This is used for variant annotation using ANNOVAR for parameter "buildver". Find more about the setting of this parameter at https://annovar.openbioinformatics.org/en/latest/user-guide/startup/. Default is "GRCh38.p13.gencode.v34.basic".
hla_ref_prefix	HLA DNA reference file in FASTA format. Default is "hla_reference_dna.fasta"
hla_ref	HLA DNA reference file path, e.g.: "/path/to/hla_reference/hla_reference_dna.fasta*".
seqtype	Reads type, "dna" or "rna". Default is "dna".
search_engine	The search engine used for MS/MS searching, comet=Comet, msgf=MS-GF+ or xtandem=X!Tandem. Default is "msgf".
search_para_file	Parameter file for the search engine used in MS/MS searching. For MS-GF+, how to set the parameter file could be find at https://msgfplus.github.io/msgfplus/MSGFPlus.html and this parameter file is used by the parameter "-conf" in MS-GF+. For Comet, how to set the parameter file could be find at http://comet-ms.sourceforge.net/parameters/ and this parameter file is used by the parameter "-P" in Comet. For X!Tandem, the parameter file setting could be found at https://www.thegpm.org/TANDEM/api/index.html
search_engine_mem	The memory limitation for MS/MS searching. Default is 60. The unit is GB.
contaminants	Contaminant protein sequence file in FASTA format. This file could be downloaded from MaxQuant website
pv_enzyme	Enzyme used for protein digestion. 0:Non enzyme, 1:Trypsin (default), 2:Trypsin (no P rule), 3:Arg-C, 4:Arg-C (no P rule), 5:Arg-N, 6:Glu-C, 7:Lys-C.This is used by PepQuery.
pv_c	The max missed cleavages, default is 2. This is used by PepQuery.
pv_tol	Precursor ion m/z tolerance, default is 10. This is used by PepQuery.
pv_tolu	The unit of --tol, ppm or Da. Default is ppm. This is used by PepQuery.
pv_itol	The error window for fragment ion, default is 0.05. This is used by PepQuery.
pv_fixmod	Fixed modification. A list of numbers separated by commas. E.g."1,2,3". Different modification is represented by different number. This is used by PepQuery. Default: "108,89,6"
pv_varmod	Variable modification. The format is the same as pv_fixmod. This is used by PepQuery. Default: "117"
netmhc_file	NetMHCpan 4.0 package in TGZ format.
neoantigen_output_prefix	The output folder name. Default is "neoflow".
pga_prefix	The prefix of PGA output file. Default is "pga".

Input manifest file format (tsv)

The input manifest file should contain the following columns:

• sample
• experiment
• wxs_file_name
• wxs_file_location
• mzml_files
  • a list of mzml files for each experiment, separated by comma
  • samples in the same experiment has the same mzml_files value
• mzml_path
  • path to mzml file for each experiment, provided as a tar file
  • samples in the same experiment has the same mzml_path value
• fusion (if fusion information is not available, 'NA' should be used)
  • path to the fusion tsv file for each sample
  • the path should match the path in the gzipped tar file (see next item)
  • all the fusion files should be tared and gzipped into a single file and provided as a command line parameter (--fusion_file)

Here is a sample manifest file.

Output

The output of NeoFlow2 is organized into folders shown below:

Below is an overview of the data in each folder:

• variant_annotation: Variant annotation result for each sample
• customized_database: Customized protein database for each sample (Label free data) or each TMT/iTRAQ experiment.
• msms_searching: MS/MS searching result. This folder contains the peptide identification files directly generated by a search engine.
• fdr_estimation: FDR estimation for peptide identification from a search engine
• pepquery: Novel peptide validation result using PepQuery for each sample.
• novel_peptide_identification: Identified novel peptides passed PepQuery validation.
• hla_type: HLA typing result for each sample.
• binding_prediction: MHC-peptide binding prediction result for each sample.
• neoantigen_prediction: This folder contains the final output files of neoantigen prediction.

For each sample, the final output file is a text file in tab delimited format. Below is the description of the columns in the file:

Columns	Description
Variant_ID	variant ID defined by neoflow
Chr	variant chromosome
Start	start position on genome
End	end position on genome
Ref	reference base
Alt	alterative base
Variant_Type	variant type annotated by ANNOVAR
Variant_Function	variant function annotated by ANNOVAR
Gene	gene ID
mRNA	mRNA ID
Neoepitope	neoepitope peptide
Variant_Start	variant start position on neoepitope peptide
Variant_End	variant end position on neoepitope peptide
AA_before	reference amino acid
AA_after	alterative amino acid
HLA_type	HLA type
netMHCpan_binding_affinity_nM	MHC-peptide binding affinity
netMHCpan_precentail_rank	MHC-peptide binding affinity rank
protein_var_evidence_pep	variant peptide