Method for idntifying of de novo peptides and feature vector generation method for identification of de novo peptides
KR20250056440A · Hanyang University · priority 2023
| Patent | KR20250056440A ↗ |
| Status | Granted |
| Priority date | 2023-10-19 |
| Filed | 2023-10-19 |
| Published | 2025-04-28 |
| Assignee | Hanyang University (printed 한양대학교 산학협력단) |
| Also published as | KR102993288B1 |
Abstract
Provided is a de novo peptide identification method for offering the high reliability. Disclosed is a method for identifying a de novo peptide, which comprises the steps of: obtaining at least one RNA sequence and at least one de novo peptide sequence for a target sample; generating at least one target sequence and at least one decoy sequence through translation of the RNA sequence; determining a matching sequence including the de novo peptide sequence among the generated target sequence and decoy sequence; inputting a feature vector for at least one matching data consisting of the de novo peptide sequence and the matching sequence into a classification model to obtain a classification score for the matching data; and searching a genome database for an RNA sequence that does not match the genome of the target sample among the RNA sequences of the target sequence determined through the classification score. The classification model is a model trained to output a high classification score when the matching sequence is the target sequence and to output a low classification score when the matching sequence is the decoy sequence.
Why it is here
Hanyang University: de novo peptide identification against translated RNA sequences, with a learned target-decoy classifier.
Inventors (2)
백은옥, 최승혁
An inventor links to an author page only where the same person is evident, never on a name alone.
Catalogued patents it cites (1)
- US11694769B2 Systems and methods for de novo peptide sequencing from data-independent acquisition using deep learning