Computational approaches to predict drug sensitivity can promote precision anticancer therapeutics. Generalizable and explainable models are of critical importance for translation to guide personalized treatment and are often overlooked in favor of prediction performance. Here, we propose PathDSP: a pathway-based model for drug sensitivity prediction that integrates chemical structure information with enrichment of cancer signaling pathways across drug-associated genes, gene expression, mutation and copy number variation data to predict drug response on the Genomics of Drug Sensitivity in Cancer dataset. Using a deep neural network, we outperform state-of-the-art deep learning models, while demonstrating good generalizability a separate dataset of the Cancer Cell Line Encyclopedia as well as provide explainable results, demonstrated through case studies that are in line with current knowledge. Additionally, our pathway-based model achieved a good performance when predicting unseen drugs and cells, with potential utility for drug development and for guiding individualized medicine.

Epstein–Barr virus (EBV) infection has long been associated with multiple sclerosis (MS), but the role of EBV in the pathogenesis of MS is not clear. Our hypothesis is that a major fraction of the expanded clones of T lymphocytes in the cerebrospinal fluid (CSF) are specific for autologous EBV-infected B cells. We obtained blood and CSF samples from eight relapsing-remitting patients in the process of diagnosis. We stimulated cells from the blood with autologous EBV-infected lymphoblastoid cell lines (LCL), EBV, varicella zoster virus, influenza, and candida and sorted the responding cells with flow cytometry after 6 d. We sequenced the RNA for T cell receptors (TCR) from CSF, unselected blood cells, and the antigen-specific cells. We used the TCR Vβ CDR3 sequences from the antigen-specific cells to assign antigen specificity to the sequences from the CSF and blood. LCL-specific cells comprised 13.0 ± 4.3% (mean ± SD) of the total reads present in CSF and 13.3 ± 7.5% of the reads present in blood. The next most abundant antigen specificity was flu, which was 4.7 ± 1.7% of the reads in the CSF and 9.3 ± 6.6% in the blood. The prominence of LCL-specific reads was even more marked in the top 1% most abundant CSF clones with statistically significant 47% mean overlap with LCL. We conclude that LCL-specific sequences form a major portion of the TCR repertoire in both CSF and blood and that expanded clones specific for LCL are present in MS CSF. This has important implications for the pathogenesis of MS.

Understanding the genetic architecture of brain structure is challenging, partly due to difficulties in designing robust, non-biased descriptors of brain morphology. Until recently, brain measures for genome-wide association studies (GWAS) consisted of traditionally expert-defined or software-derived image-derived phenotypes (IDPs) that are often based on theoretical preconceptions or computed from limited amounts of data. Here, we present an approach to derive brain imaging phenotypes using unsupervised deep representation learning. We train a 3-D convolutional autoencoder model with reconstruction loss on 6130 UK Biobank (UKBB) participants’ T1 or T2-FLAIR (T2) brain MRIs to create a 128-dimensional representation known as Unsupervised Deep learning derived Imaging Phenotypes (UDIPs). GWAS of these UDIPs in held-out UKBB subjects (n = 22,880 discovery and n = 12,359/11,265 replication cohorts for T1/T2) identified 9457 significant SNPs organized into 97 independent genetic loci of which 60 loci were replicated. Twenty-six loci were not reported in earlier T1 and T2 IDP-based UK Biobank GWAS. We developed a perturbation-based decoder interpretation approach to show that these loci are associated with UDIPs mapped to multiple relevant brain regions. Our results established unsupervised deep learning can derive robust, unbiased, heritable, and interpretable brain imaging phenotypes.

Drug combination therapy has become a promising therapeutic strategy for cancer treatment. While high-throughput drug combination screening is effective for identifying synergistic drug combinations, measuring all possible combinations is impractical due to the vast space of therapeutic agents and cell lines. In this study, we propose a biologically-motivated deep learning approach to identify pathway-level features from drug and cell lines' molecular data for predicting drug synergy and quantifying the interactions in synergistic drug pairs. This method obtained an MSE of 70.6 ± 6.4, significantly surpassing previous approaches while providing potential candidate pathways to explain the prediction. We further demonstrate that drug combinations tend to be more synergistic when their top contributing pathways are closer to each other on a protein interaction network, suggesting a potential strategy for combination therapy with topologically interacting pathways. Our computational approach can thus be utilized both for prescreening of potential drug combinations and for designing new combinations based on proximity of pathways associated with drug targets and cell lines. Implications: Our computational framework may be translated in the future to clinical scenarios where synergistic drugs are tailored to the patient and additionally, drug development could benefit from designing drugs that target topologically close pathways.

Journal of Medical Internet Research - International Scientific Journal for Medical Research, Information and Communication on the Internet

Traumatic brain injury (TBI) results in a cascade of cellular responses, which produce neuroinflammation, partly due to the activation of microglia. Accurate identification of microglial populations is key to understanding therapeutic approaches that modify microglial responses to TBI and improve long-term outcome measures. Notably, previous studies often utilized an outdated convention to describe microglial phenotypes. We conducted a temporal analysis of the response to controlled cortical impact (CCI) in rat microglia between ipsilateral and contralateral hemispheres across seven time points, identified microglia through expression of activation markers including CD45, CD11b/c, and p2y12 receptor and evaluated their activation state using additional markers of CD32, CD86, RT1B, CD200R, and CD163. We identified unique sub-populations of microglial cells that express individual or combination of activation markers across time points. We further portrayed how the size of these sub-populations changes through time, corresponding to stages in TBI response. We described longitudinal changes in microglial population after CCI in two different locations using activation markers, showing clear separation into cellular sub-populations that feature different temporal patterns of markers after injury. These changes may aid in understanding the symptomatic progression following TBI and help define microglial subpopulations beyond the outdated M1/M2 paradigm.