The invention relates to a
tumor antigen prediction method based on a whole
transcriptome. The
tumor antigen prediction method includes the steps: respectively performing
protein generation and
peptide fragment interception of tumor-associated
antigen, detection of tumor somatic
mutation and corresponding
mutant peptide fragment interception, generation of tumor-specific novel transcript and
peptide fragment interception, and
gene fusion detection and
fusion peptide fragment interception in tumor tissues, according to the whole
transcriptome sequencing data of tumor tissues and corresponding adjacent para-
carcinoma tissues, obtaining tumor-associated
antigen, tumor somatic
mutation, tumor novel transcript, tumor-specific
peptide fragment of
gene fusion, calculating affinity of the obtainedtumor-specific
peptide fragment and the HLA molecule and the amount of expression in each transcript, and based on the affinity value of the tumor-specific
peptide fragment and the amount of expression TPM (transcripts per million) value, evaluating the level of the candidate
tumor antigen. The invention also provides an application of the same. The tumor
antigen prediction method based on a whole
transcriptome and the application of the same are conductive to accurate calculation of the tumor antigen load, evaluation of immunological therapy effect, and tumor vaccine design of the later stage of service.