Thursday , August 11 2022

A new method identifies cancer-related genes from artificial intelligence


Posted 02/09/2019 11:41:31CET

Patients with breast tumors contain a protein

Barcelona, ​​February 19 (EUROPA PRESS) –

The National Center for Supercomputing (BCC) at the Barcelona SuperCountry Center has created a new computational system based on machine learning tools (& # 39; machine learning & # 39;) that help identify new genes related to cancer. It can be found in 36 relationship.

Study published in & # 39; Nature Communications & # 39; has proven biological method and its results. In relation to OMIC data in large quantities, & # 39; Machine Learning & # 39; s use a computational prototype integrated cell or & # 39; Isel & # 39; Help us detect new biomarkers and improve forecasts.

New scientist NASAA Przlaj said that genetically-associated cancer changes do not allow any other change to be altered. Tuesday explained the center.

"This discovery biological data analysis highlights the importance of mutual approach to facilitate the comparative analysis of all cells", added the researcher.

In other experiments to prove this, patients with high protein breast hybrid proteins – MRLP3 – were in a low condition if not previously associated with cancer.

This discovery is an example of the way to detect new biomarker genes relevant to the dramatic and predictability of the diagnosis.

36 years old

It is used to repair cells from four cancers such as cancer, prostate, lung colon, and colon. It is effective in finding new genes associated with these diseases.

This system has identified only 63 genes. It is found that at least 36 of the biological validation process leads to irregular growth of cells.

The prototype designed by the BSC through a fusion technique, first proposed for the addition of prototine protein, genes and genetic interactive networks, combines three Tissue Specific Molecular Intrusion Networks. Use recently for data integration

Source link