From: Neoantigen-targeted TCR-engineered T cell immunotherapy: current advances and challenges
Prediction tool Pipeline | Time | Input data type | Workflow and Features | HLA allele | MHC-peptide | Validation in vivo or in vitro |
---|---|---|---|---|---|---|
INTEGRATE-neo | 2016 | SV | Starting from the original FASTA/FASTAQ file, the main steps included prediction of fusion gene peptides, prediction of HLA alleles, and prediction of neoantigens | HLAminer | NetMHC4.0 | No |
[26] | ||||||
TSNAD/ TSNAD v2.0 | 2017/2021 | SNV, indel / SNV, indel, fusion | TSNAD is a one-stop software solution for predicting neoantigens from WES/WGS data of tumor-normal pairs. The new version adds RNA-seq data analysis, supports two versions of the reference genome (GRCh37 and GRCh38), uses DeepHLApan instead of NetMHCpan, and provides Web services and Docker installation methods | OptiType | DeepHLApan | No |
MuPeXI | 2017 | SNV, indel | MuPeXI acceptes WES/WGS and RNA-seq sequencing data as source data. MuPeXI measures the immunogenicity of new antigens based on quantitative scores of important features of new peptides, prioritizing predicted peptides based on affinity of mutant and normal peptides, allele frequencies of mutant peptides, and gene expression levels | OptiType | NetMHCpan4.0 | |
[29] | In vivo validation [33] | |||||
In vitro validation [34] | ||||||
CloudNeo | 2017 | SNV, indel | CloudNeo is a cloud-based workflow for identifying tumor neoantigens in patients, which is a common workflow language implementation of HLA typing using Polysolver or HLAminer combined with custom scripts for mutant peptide identification and NetMHCpan for neoantigen prediction | HLAminer/Polysolver | NetMHCpan3.0 | No |
[35] | ||||||
Timiner | 2017 | SNV, indel | TIminer integrates a suite of bioinformatics tools to analyze RNA-seq data and somatic DNA mutations from a single sample, including: 1) HLA genotyping from NGS data; 2) Using mutation data and HLA types to predict tumor neoantigens; 3) Identification of tumor infiltrating immune cells from RNA-seq data; 4) Analysis of tumor immunity profile from expression data | Optitype | NetMHCpan3.0 | No |
[36] | ||||||
pTuneos | 2019 | SNV, indel | pTuneos presents a computational framework for tumor de novo antigen sequencing and screening to effectively predict neoantigens from high-throughput sequencing data and evaluate and rank their true immunogenicity. Moreover, pTunes architecture implements multi-thread processing, which improves the speed of high-throughput sequencing data processing | OptiType | NetMHCpan4.0 | No |
[37] | ||||||
Neo-Fusion | 2019 | SV | Neo-Fusion utilizes two separate ion database searches to identify the two halves of each spliced peptide and infer the complete spliced sequence. The feature of this tool is that it allows the identification of spliced peptides without the restriction of peptide length, providing a widely applicable tool for the identification of spliced peptides | OptiType | NetMHCpan | No |
[38] | ||||||
NeoPredPipe | 2019 | SNV, indel | Allow users to process neoantigens predicted from single or multi-region vcf files using ANNOVAR and NetMHCpan | POLYSOLVER | NetMHCpan | No |
[39] | ||||||
NeoFuse | 2020 | Fusion | NeoFuse is a computational pipeline for the prediction of fusion neoantigens from tumor RNA-seq data, which is available as Singularity and Docker images to simplify installation and analysis | OptiType | MHCflurry | No |
[40] | ||||||
NeoFlow | 2020 | SNV, indel | NeoFlow consists of four functional modules: (1)Â variation annotation and construction of sample personalized protein database; (2)Â Identification of peptides based on mass spectrometry data; (3)Â HLA type prediction based on WGS data or WES data; (4)Â Neoantigen prediction. This process can also be used to analyze data from immune polypeptide groups | - | - | No |
[41] | ||||||
OpenVax | 2020 | SNV, indel | OpenVax is a computational workflow for identifying somatic mutations, predicting neoantigens, and developing personalized cancer vaccines. OpenVax is also an end-to-end workflow that starts with raw DNA and RNA FASTQ data, generates a mutant containing peptide, and finally outputs a specified length of synthetic growth peptide containing the mutant peptide segment | - | Different options | NCT02721043 |
[42] | NCT03223103 | |||||
NCT03359239 | ||||||
pVACtools | 2020 | SNV, indel,fusion | pVACtools is an integrated computing tool pipeline composed of five main parts, including pVACseq, pVACbind, pVACfuse, pVACvector, and pVACviz, which is highly modularized and is divided into flexible components that can be run independently. pVAC-tools can support the identification of mutant peptides from from different sources, including missense, frameshift mutations, insertions-deletions, and gene fusions. The predicted peptides were prioritized by integrating different data, including the expression of the mutant allele, the affinity of the binding peptide, and the type of clone. pVACtools has been used to predict and develop cancer vaccines for immunotherapy research in clinical trials | HLAminer/Athlates | 8 MHC Class I algorithms: NetMHCpan, NetMHC, NetMHCcons, PickPocket, SMM, SMMPMBEC, MHCflurry, and MHCnuggets; | NCT00683670 [43] |
[44] | 4 MHC Class II algorithms: NetMHCIIpan, SMMalign, NNalign, and MHCnuggets | |||||
In vivo validation [49] | ||||||
ASNEO | 2020 | SV | ASNEO is an integrated computational pipeline that analyzes RNA-seq data to identify neoantigens generated by personalized alternative splicing | OptiType | NetMHCpan4.0 | No |
[50] | ||||||
Neoepiscope | 2020 | SNV, indel | Neoepiscope is used to predict epitopes from DNA-seq data, which can incorporate germline context and address variant phasing for SNVs and indels. Neoepiscope framework is sufficiently flexible to accommodate numerous variant types, nonsense-mediated decay products and epitope prediction across different genomes | Optitype,PHLAT | MHCflurry, MHCnuggets or independently install NetMHCpan and NetMHCIIpan | No |
[51] | ||||||
neoANT-HILL | 2020 | SNV, indel | neoANT-HILL integrates multiple immunogenomic analyses combined with quantification of immune cells infiltrating tumors and considers the use of RNA-Seq data to identify potential de novo antigens. In particular, neoANT-HILL is a user-friendly software tool with a graphical interface that can be used by users without programming skills | Optitype | Employs seven binding prediction algorithms: NetMHC, NetMHCpan 4.0, NetMHCcons, NetMHCstabpan, PickPocket, SMM and SMMPMBEC, MHCflurry | No |
[52] | ||||||
TruNeo | 2020 | SNV, indel, | TruNeo pipeline requires to input raw DNA sequencing and RNA-seq data. The prediction steps include annotating somatic mutation information, obtaining HLA genotype and gene expression information; Candidate neoantigens were predicted according to the affinity of peptide binding to MHC. The candidate neoantigens were scored by integrating the information of multiple neoantigen presentation processes, and the high confidence neoantigens were finally screened and output | Polysolver and BWA-HLA | NetMHCpan3.0 | In vitro validation [53] |
[54] | ||||||
ProGeo-Neo v2.0/ProGeo-Neo | 2022 | SNV, indel, fusion | ProGeo-neo v2.0 is a proteogenomics-based neoantigen prediction pipeline, a one-stop software, which is divided into five modules: WES/WGS data processing; RNA-seq data processing; The construction of customized mutant protein sequence database and the identification of MS data database; Prediction of neoantigens; Computational screening of neoantigens | OptiType | NetMHCpanI 4.1 | No |
NetMHCpanII 4.0 | ||||||
PGNneo | 2023 | Noncoding regions | PGNneo is a proteogenomics-based pipeline developed for predicting neoantigens from non-coding regions. The pipeline consists of four modules: (1)Â non-coding somatic variant calling and HLA typing; (2)Â peptide extraction and customized database construction; (3) identification of variant peptides; (4)Â selection of candidate neoantigens | OptiType | NetMHCpan 4.1 | No |
[57] |