Method for de novo sequencing of amino acid sequences
RU2843981C1 · Institute of Biomedical Chemistry
| Patent | RU2843981C1 ↗ |
| Filed | 2024-12-03 |
| Published (granted) | 2025-07-22 |
| Assignee | Institute of Biomedical Chemistry (printed Федеральное государственное бюджетное научное учреждение “Научно-исследовательский институт биомедицинской химии имени В.Н. Ореховича” (ИБМХ)) |
| Claims | PowerNovo |
Abstract
FIELD: bioinformatics. SUBSTANCE: described is a method for de novo sequencing of amino acid sequences. In this method, the studied spectrum electronic data are converted to the vector form; finding the next symbol in the identified amino acid sequence occurrence probabilities distribution, accessing the transformer trained neural network; selecting a predetermined number of symbols sequence versions with maximum probabilities; repeating this process, resulting in decoded versions of amino acid sequences; evaluating the hypotheses of the decoded versions existence, referring to the trained evaluating neural network; storing the electronic data of those decoded amino acid sequences, the evaluation of the existence hypotheses of which meet predetermined conditions; combining stored amino acid sequences electronic data into continuous overlapping fragments; comparing the electronic data of each said fragment with electronic data of amino acid sequences of real proteins from a pre-created database of known and predicted proteins in order to determine the most probable proteins in the studied spectrum; eliminating the ambiguity of comparing amino acid sequences with proteins by selecting such proteins for which maximum intersection is achieved on all amino acid sequences found from the studied spectrum data. EFFECT: use of the invention provides a technical result in the form of a wider field of application, increase in the number and accuracy of identifying proteins and peptides in complex samples, including those of biological origin. 2 cl, 1 dwg, 2 ex
Why it is here
Institute of Biomedical Chemistry (Moscow): autoregressive de novo sequencing, predicting each next residue from the vectorised spectrum; the inventors are PowerNovo’s authors.
Inventors (5)
Denis V. Petrovskiy, Kirill S. Nikolsky, Liudmila I. Kulikova, Vladimir R. Rudnev, Anna L. Kaysheva
An inventor links to an author page only where the same person is evident, never on a name alone.
Catalogued patents it cites (1)
- US7783429B2 Peptide sequencing from peptide fragmentation mass spectra