Efficient Algorithms For The Analysis Of Protein Sequences And Detecting Prostate Cancer Using Machine Learning Techniques

Abstract

Prostate Cancer (PC) diagnosis has an issue with over diagnosis, which causes newlineharmful side effects from unneeded therapy. It is demonstrated that multi-parametric newlinemagnetic resonance imaging greatly raises the standard of biopsies. Reduce over newlinediagnosis as much as possible in order to minimize its detrimental effects on healthy newlineindividuals. This research aims to investigate the use of MRI as an input in computeraided diagnostics utilizing deep learning techniques. In order to address a range of newlinediagnosis-related problems, including the classification, and early detection, deep newlinelearning-based algorithms are applied. newlineSince the traditional methods face difficulties in detecting prostate cancer, in phase, newlinewe used a prostate cancer gene-related approach, and for that, Protein-to-Protein newlineInteraction (PPI) network is utilized. Here, a PPI network with weights was built newlineusing PPI information got from the Search Tool for the Recovery of Relating newlineGenes/Proteins (STRING) database. A total of 18 prostate-associated genes were newlinemined from the database via the Kyoto Encyclopedia of Genes and Genomes (KEGG) newlinepathway. The direct (shortest) path/route among 18 genes was then discovered with newlinethe Non-Direct, Non-Reverse Shortest Path (NDNRSP) method. Discovering the newlineshortest route/path among genes associated with prostate cancer was typically quite newlineeasy using the traditional Ford-Fulkerson method, although the time taken was long newlinewhen the PPI network comprised additional number of genes. The NDNRSP newlinetechnique was developed to save computing time by deleting the incorrect gene newlineconnection route. newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced