Short-length peptides contact map prediction using Convolution Neural Networks

A. Maminov

АБВ
АБВ
АБВ

Обычная версия сайта

Priority areas

by year

Subject

News

August 25, 2026

Scientists Develop Algorithm for More Reliable Processors in Data Centres

Researchers from HSE MIEM and Samara University have developed the LRF-3D algorithm to automatically bypass idle nodes in three-dimensional networks-on-chip. Thanks to its hierarchical architecture, the algorithm outperforms existing solutions in both speed and path accuracy, improving processor reliability for use in data centres, supercomputers, and AI computing. The source code and test results are publicly available.

August 24, 2026

Researchers Develop Method for Direct Generation of Regulatory DNA

Researchers at HSE University have developed a model for generating promoters and enhancers—DNA sequences that regulate gene activity. The model works directly with DNA nucleotides, without first transforming them into a continuous numerical representation. This solution could be useful for applications in synthetic biology and gene therapy. The study results were presented at the ICLR 2026 Workshop ‘Generative AI in Genomics (Gen^2): Barriers and Frontiers.’

August 21, 2026

Social Integration: At the Crossroads of Knowledge and Values

The International Laboratory for Social Integration Research (ILSIR) at HSE University studies the challenges faced by vulnerable groups and explores ways to help them participate fully in everyday life. To develop effective solutions, the laboratory’s researchers combine cutting-edge methods with practical fieldwork. In this interview with the HSE News Service, Laboratory Head Elena Iarskaia-Smirnova discusses the laboratory’s work.

Have you spotted a typo?
Highlight it, click Ctrl+Enter and send us a message. Thank you for your help!

Publications

?

Short-length peptides contact map prediction using Convolution Neural Networks

Maminov A.

Inthisarticle, it isconsideredanapproachforpredictingthecontactmatrix(contactmap) for short-lengthpeptides. Contactmatrixis two-dimensional representationof theprotein. Itcan beusedfortertiarystructurereconstructionorforstartingapproximationinenergyminimization models.Forthiswork,peptideswithachainlengthfrom15upto30werechosentotestthemodel andsimplifythecalculations. Convolutionalneuralnetworks(CNNs)wereusedasaprediction toolaccordingtothefactthatthefeaturespaceofeachpeptideispresentedasatwo-dimensional matrix. SCRATCHtoolwasusedtogeneratethesecondarystructure,solventaccessibility,and profilematrix(PSSM) foreachpeptide. CNNwas implementedinthePythonprogramming languageusingtheKeras library. ToworkwiththecommonPDB-format,whichpresents the structureinformationofproteins,theBioPythonmodulewasused.Asaresult,training,validation andtestsamplesweregenerated, themultilayermulti-outputconvolutionalneuralnetworkwas constructed,whichwastrainedandvalidated.Theexperimentswereconductedonatestsample topredictthecontactmatrixandcompareitwithnativeone.Toassessthequalityofprediction, conjunctionmatricesfor thethresholdof8and12𝐴wereformed, themetricsF1-score, recall andprecisionwerecalculated. According toF1-score,wecanobserve, that evenwithsmall neuralnetworkwecanachevequitegoodresults.AtthefinalstepFT-COMARtoolwasusedto reconstruct tertiarystructureoftheproteinsfromitscontactmatrix. Theresultsshows, thatfor reconstructedstructuresfrom12threshholdcontactmatrix,RMSDmetricisbetter

Language: English

Keywords: Folding

In book

Proceedings of Science, volume 429. The 6th International Workshop on Deep Learning in Computational Physics

[б.и.], 2023.