izpis_h1_title_alt

Self-supervised learning of Cox regression for latent gene set representation
ID Špendl, Martin (Author), ID Zupan, Blaž (Mentor) More about this mentor... This link opens in a new window

.pdfPDF - Presentation file, Download (1009,55 KB)
MD5: E9C0A779C52707C71E63A926E9429019

Abstract
Enhanced understanding of disease mechanisms on a molecular level leads to more effective treatment decisions. The high-dimensional nature of such data requires dimensionality reduction techniques to extract important patterns. To improve the interpretation and relevance of latent representations, domain knowledge is introduced during modeling by influencing data preprocessing or model architecture, while domain-inspired loss functions are scarcely explored. We propose a novel autoencoder loss function for modeling mRNA concentration based on the first-order differential equation of mRNA dynamics. We decompose the concentration into transcription (synthesis) and mRNA decay and reformulate the problem as survival analysis. By extending the definition of CoxPH partial likelihood, we perform gradient descent through both risk and survival time, which achieves the correct interpretation of both processes. Representations of clinical samples and cell-line data show increased performance on clustering and drug response prediction tasks compared to standard variational autoencoders.

Language:English
Keywords:variational autoencoders, mRNA dynamics, dimensionality reduction, survival analysis
Work type:Master's thesis/paper
Organization:FRI - Faculty of Computer and Information Science
Year:2024
PID:20.500.12556/RUL-166125 This link opens in a new window
Publication date in RUL:20.12.2024
Views:19
Downloads:2
Metadata:XML DC-XML DC-RDF
:
Copy citation
Share:Bookmark and Share

Secondary language

Language:Slovenian
Title:Samospodbujevano učenje Coxovega regresijskega modela za predstavitev genskih skupin v latentnem prostoru
Abstract:
Izboljšano razumevanje mehanizmov bolezni na molekulski ravni vodi do učinkovitejše terapije. Visokodimenzionalna narava takšnih podatkov zahteva uporabo tehnik zmanjšanja dimenzionalnosti, za pridobivanje ključnih vzorcev. Za izboljšano interpretacijo in relevantnost predstavitev v proces modeliranja lahko vključimo domensko predznanje, ali v pred obdelavi podatkov ali z izbiro arhitekture modela, medtem ko so vpliv domenskega znanja na cenilne funkcije še ni raziskan. Predlagamo novo cenilno funkcijo samokodirnikov, namenjeno modeliranju koncentracije mRNA, ki temelji na diferencialni enačbi prvega reda dinamike mRNA. Koncentracijo razdelimo na transkripcijo (sintezo) in razgradnjo mRNA ter problem reformuliramo kot analizo preživetja. Z razširitvijo definicije verjetnostne porazdelitve CoxPH modela dosežemo izvedbo gradientnega spusta skozi tveganje in čas preživetja, kar omogoča pravilno interpretacijo obeh bioloških procesov. Predstavitve kliničnih vzorcev in celičnih linij dosegajo boljše rezultate pri nalogah gručenja in napovedi odziva na zdravila v primerjavi s standardnimi variacijskimi samokodirniki.

Keywords:variacijski samokodirnik, dinamika mRNA, zmanjševanje dimenzij, analiza preživetja

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Back