Bayesian Machine Learning Enables Identification of Transcriptional Network Disruptions Associated with Drug-Resistant Prostate Cancer

Abstract

Survival rates of patients with metastatic castration-resistant prostate cancer (mCRPC) are low due to lack of response or acquired resistance to available therapies, such as abiraterone (Abi). A better understanding of the underlying molecular mechanisms is needed to identify effective targets to overcome resistance. Given the complexity of the transcriptional dynamics in cells, differential gene expression analysis of bulk transcriptomics data cannot provide sufficient detailed insights into resistance mechanisms. Incorporating network structures could overcome this limitation to provide a global and functional perspective of Abi-resistance in mCRPC. Here, we developed TraRe, a computational method using sparse Bayesian models to examine phenotypically-driven transcriptional mechanistic differences at three distinct levels: transcriptional networks, specific regulons, and individual transcription factors. TraRe was applied to transcriptomic data from 46 mCRPC patients with Abi-response clinical data and uncovered abrogated immune response transcriptional modules that showed strong differential regulation in Abi-responsive compared to Abi-resistant patients. These modules were replicated in an independent mCRPC study. Further, key rewiring predictions and their associated transcription factors were experimentally validated in two prostate cancer cell lines with different Abi resistance features. Among them, ELK3, MXD1, and MYB played a differential role in cell survival in Abi-sensitive and Abi-resistant cells. Moreover, ELK3 regulated cell migration capacity, which could have a direct impact on mCRPC. Collectively, these findings shed light on the underlying transcriptional mechanisms driving abiraterone response, demonstrating that TraRe is a promising tool for generating novel hypotheses based on identified transcriptional network disruptions.

NetActivity enhances transcriptional signals by combining gene expression into robust gene set activity scores through interpretable autoencoders

“…the relationship between the genes is similar, although the gene expression can be very different. For other gene sets, their representation varies, suggesting a different relationship between the genes in normal and cancer tissues, as reported in the literature ( 50 ). Overall, NetActivity can learn data specificities while preserving the coherent biological information of each dataset.…”

Developments and future prospects of personalized medicine in head and neck squamous cell carcinoma diagnoses and treatments

“…ML initiatives have developed prediction models employing various techniques like penalized logistical regression, artificial neural networks (ANNs), Bayesian networks (BNs), decision trees (DTs), and support vector machines (SVMs). Studies have highlighted SVM's impressive accuracy in classifying survival and recurrence among patients with breast cancer, oral cancer, and cervical cancer 189–195 . Segmentation plays a vital role in radiomics analysis, particularly in radiotherapy treatment 196 .…”

“…Studies have highlighted SVM's impressive accuracy in classifying survival and recurrence among patients with breast cancer, oral cancer, and cervical cancer. 189 , 190 , 191 , 192 , 193 , 194 , 195 Segmentation plays a vital role in radiomics analysis, particularly in radiotherapy treatment. 196 The planning target volume (PTV), which includes the gross tumor volume (GTV) and allows for daily setup uncertainties and organ movements during treatment, accounts for potential microscopic tumor spread.…”

Emerging role of MYB transcription factors in cancer drug resistance

“…MYB interacts directly with AR, leading to ligand-independent AR activation and the establishment of problematic CRPC [109] . The combination of the semi-synthetic steroidal cytochrome P450 17A1 (CYP17A1) inhibitor abiraterone with prednisone is a first-line therapy of CRPC [110] . Abirateronesensitive prostate cancer cells revealed much higher MYB levels than abiraterone-resistant cells, and the reduction of MYB signaling seemed responsible for the development of abiraterone resistance [111] .…”

“…The combination of the semi-synthetic steroidal cytochrome P450 17A1 (CYP17A1) inhibitor abiraterone with prednisone is a first-line therapy of CRPC [ 110 ] . Abiraterone-sensitive prostate cancer cells revealed much higher MYB levels than abiraterone-resistant cells, and the reduction of MYB signaling seemed responsible for the development of abiraterone resistance [ 111 ] . The observations that antiandrogen therapy upregulates MYB expression and abiraterone activity depends on active MYB signaling might solve clinical problems of current CRPC therapy.…”