Technology for reduction feature space for classification immunosignature data

Tracking #: 555-1535

Authors:

NameORCID
Владимир АндрющенкоORCID logo https://orcid.org/0000-0002-9757-0733
Alexander KoshechkinORCID logo https://orcid.org/0000-0002-9751-1565
Olga RomanovichORCID logo https://orcid.org/0000-0001-5698-991X
Daniel StamateORCID logo https://orcid.org/0000-0001-8565-6890
Alexander ZamyatinORCID logo https://orcid.org/0000-0002-1416-7472


Responsible editor: 

Núria Queralt-Rosinach

Submission Type: 

Position Paper

Abstract: 

Random sequences of peptides in a microchip make it possible to generate specific immunosignatures that can diagnose various diseases. A large number of features does not allow for the quick and efficient analysis of such data. In this study, we propose technology to reduce feature space using various methods. The proposed technology makes it possible with minimal computational costs to ensure the accuracy and reliability of the classification of immunosignature data. The technology was tested on samples formed from a set of real data with the introduction of noise at various levels. The efficiency of the proposed technology on all test samples with various classifiers used for further data analysis is shown. .

Manuscript: 

Tags: 

  • Under Review

Data repository URLs: 

https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE52580

Date of Submission: 

Tuesday, January 29, 2019