A biophysically inspired model of transcriptional regulation in hormone signalling

Prostate cancer remains one of the leading causes of cancer-related deaths in men, with androgen receptor (AR) functioning as a key transcription factor (TF) driving disease progression. While AR's role in transcription is well established, the other TFs and regulatory mechanisms that govern AR-mediated transcription are not fully understood. To address this, high-throughput STARR-seq assays have measured the transcriptional activity of thousands of putative AR binding sites in LNCaP cells following androgen treatment. Using this dataset, we have developed a biophysically motivated model of transcriptional regulation based on a known set of TFs expressed in LNCaP cells. The model's parameters are directly interpretable as effective interactions between TFs and RNA polymerase, offering insights into AR-mediated transcription. Our results identify TFs that regulate transcription independently of AR, as well as those that collaborate with AR in response to hormone signaling. We also observe a hierarchical spatial organization within enhancer sequences, consistent with the formation of larger TF complexes. Lastly, our model is able to predict the effects of point mutations within functional enhancer sequences, providing a deeper understanding of how mutations affect transcription.