The emergence of large-scale omics data and foundational models has renewed efforts in the field of cancer drug response prediction (DRP). Despite recent progress, challenges such as limited tumor heterogeneity in standard cell lines and inconsistencies in experimental protocols across studies persist. However, these challenges also open significant opportunities for innovation. The complex nature of drug responses, influenced by variations in new patients and new drugs, presents a critical area for advancing validation approaches that traditional machine learning approaches often overlook. This review provides a comprehensive overview of the current state of DRP using advanced machine-learning models, discussing data sources, model designs, and evaluation methods. We introduce a unified framework for testing these models with a focus on clinically relevant metrics. By evaluating a range of foundational and deep-learning models within this framework, we identify performance gaps and propose concrete strategies to advance these computational models for reliable use in personalized cancer treatment, thereby unlocking their full clinical potential.

The success of machine learning models relies heavily on effectively representing high-dimensional data. However, ensuring data representations capture human-understandable concepts remains difficult, often requiring the incorporation of prior knowledge and decomposition of data into multiple subspaces. Traditional linear methods fall short in modeling more than one space, while more expressive deep learning approaches lack interpretability. Here, we introduce Supervised Independent Subspace Principal Component Analysis ($\texttt{sisPCA}$), a PCA extension designed for multi-subspace learning. Leveraging the Hilbert-Schmidt Independence Criterion (HSIC), $\texttt{sisPCA}$ incorporates supervision and simultaneously ensures subspace disentanglement. We demonstrate $\texttt{sisPCA}$'s connections with autoencoders and regularized linear regression and showcase its ability to identify and separate hidden data structures through extensive applications, including breast cancer diagnosis from image features, learning aging-associated DNA methylation changes, and single-cell analysis of malaria infection. Our results reveal distinct functional pathways associated with malaria colonization, underscoring the essentiality of explainable representation in high-dimensional data analysis.

MOTIVATION: The human leukocyte antigen (HLA) locus plays a critical role in tissue compatibility and regulates the host response to many diseases, including cancers and autoimmune di3orders. Recent improvements in the quality and accessibility of next-generation sequencing have made HLA typing from standard short-read data practical. However, this task remains challenging given the high level of polymorphism and homology between HLA genes. HLA typing from RNA sequencing is further complicated by post-transcriptional modifications and bias due to amplification. RESULTS: Here, we present arcasHLA: a fast and accurate in silico tool that infers HLA genotypes from RNA-sequencing data. Our tool outperforms established tools on the gold-standard benchmark dataset for HLA typing in terms of both accuracy and speed, with an accuracy rate of 100% at two-field resolution for Class I genes, and over 99.7% for Class II. Furthermore, we evaluate the performance of our tool on a new biological dataset of 447 single-end total RNA samples from nasopharyngeal swabs, and establish the applicability of arcasHLA in metatranscriptome studies. AVAILABILITY AND IMPLEMENTATION: arcasHLA is available at https://github.com/RabadanLab/arcasHLA. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Analyzing large-scale, multi-experiment studies requires scientists to test each experimental outcome for statistical significance and then assess the results as a whole. We present Black Box FDR (BB-FDR), an empirical-Bayes method for analyzing multi-experiment studies when many covariates are gathered per experiment. BB-FDR learns a series of black box predictive models to boost power and control the false discovery rate (FDR) at two stages of study analysis. In Stage 1, it uses a deep neural network prior to report which experiments yielded significant outcomes. In Stage 2, a separate black box model of each covariate is used to select features that have significant predictive power across all experiments. In benchmarks, BB-FDR outperforms competing state-of-the-art methods in both stages of analysis. We apply BB-FDR to two real studies on cancer drug efficacy. For both studies, BB-FDR increases the proportion of significant outcomes discovered and selects variables that reveal key genomic drivers of drug sensitivity and resistance in cancer.

The recent explosion of genomic data has underscored the need for interpretable and comprehensive analyses that can capture complex phylogenetic relationships within and across species. Recombination, reassortment and horizontal gene transfer constitute examples of pervasive biological phenomena that cannot be captured by tree-like representations. Starting from hundreds of genomes, we are interested in the reconstruction of potential evolutionary histories leading to the observed data. Ancestral recombination graphs represent potential histories that explicitly accommodate recombination and mutation events across orthologous genomes. However, they are computationally costly to reconstruct, usually being infeasible for more than few tens of genomes. Recently, Topological Data Analysis (TDA) methods have been proposed as robust and scalable methods that can capture the genetic scale and frequency of recombination. We build upon previous TDA developments for detecting and quantifying recombination, and present a novel framework that can be applied to hundreds of genomes and can be interpreted in terms of minimal histories of mutation and recombination events, quantifying the scales and identifying the genomic locations of recombinations. We implement this framework in a software package, called TARGet, and apply it to several examples, including small migration between different populations, human recombination, and horizontal evolution in finches inhabiting the Galápagos Islands.

BACKGROUND: Patterns of disease incidence can identify new risk factors for the disease or provide insight into the etiology. For example, allergies and infectious diseases have been shown to follow periodic temporal patterns due to seasonal changes in environmental or infectious agents. Previous work searching for seasonal or other temporal patterns in disease diagnosis rates has been limited both in the scope of the diseases examined and in the ability to distinguish unexpected seasonal patterns. Electronic Health Records (EHR) compile extensive longitudinal clinical information, constituting a unique source for discovery of trends in occurrence of disease. However, the data suffer from inherent biases that preclude an identification of temporal trends. METHODS: Motivated by observation of the biases in this data source, we developed a method (Lomb-Scargle periodograms in detrended data, LSP-detrend) to find periodic patterns by adjusting the temporal information for broad trends in incidence, as well as seasonal changes in total hospitalizations. LSP-detrend can sensitively uncover periodic temporal patterns in the corrected data and identify the significance of the trend. We apply LSP-detrend to a compilation of records from 1.5 million patients encoded by ICD-9-CM (International Classification of Diseases, Ninth Revision, Clinical Modification), including 2,805 disorders with more than 500 occurrences across a 12 year period, recorded from 1.5 million patients. RESULTS AND CONCLUSIONS: Although EHR data, and ICD-9 coded records in particular, were not created with the intention of aggregated use for research, these data can in fact be mined for periodic patterns in incidence of disease, if confounders are properly removed. Of all diagnoses, around 10% are identified as seasonal by LSP-detrend, including many known phenomena. We robustly reproduce previous findings, even for relatively rare diseases. For instance, Kawasaki disease, a rare childhood disease that has been associated with weather patterns, is detected as strongly linked with winter months. Among the novel results, we find a bi-annual increase in exacerbations of myasthenia gravis, a potentially life threatening complication of an autoimmune disease. We dissect the causes of this seasonal incidence and propose that factors predisposing patients to this event vary through the year.

BACKGROUND: Drug-drug interactions (DDIs) are responsible for many serious adverse events; their detection is crucial for patient safety but is very challenging. Currently, the US Food and Drug Administration and pharmaceutical companies are showing great interest in the development of improved tools for identifying DDIs. METHODS: We present a new methodology applicable on a large scale that identifies novel DDIs based on molecular structural similarity to drugs involved in established DDIs. The underlying assumption is that if drug A and drug B interact to produce a specific biological effect, then drugs similar to drug A (or drug B) are likely to interact with drug B (or drug A) to produce the same effect. DrugBank was used as a resource for collecting 9454 established DDIs. The structural similarity of all pairs of drugs in DrugBank was computed to identify DDI candidates. RESULTS: The methodology was evaluated using as a gold standard the interactions retrieved from the initial DrugBank database. Results demonstrated an overall sensitivity of 0.68, specificity of 0.96, and precision of 0.26. Additionally, the methodology was also evaluated in an independent test using the Micromedex/Drugdex database. CONCLUSION: The proposed methodology is simple, efficient, allows the investigation of large numbers of drugs, and helps highlight the etiology of DDI. A database of 58 403 predicted DDIs with structural evidence is provided as an open resource for investigators seeking to analyze DDIs.

BACKGROUND: Adverse drug events (ADE) cause considerable harm to patients, and consequently their detection is critical for patient safety. The US Food and Drug Administration maintains an adverse event reporting system (AERS) to facilitate the detection of ADE in drugs. Various data mining approaches have been developed that use AERS to detect signals identifying associations between drugs and ADE. The signals must then be monitored further by domain experts, which is a time-consuming task. OBJECTIVE: To develop a new methodology that combines existing data mining algorithms with chemical information by analysis of molecular fingerprints to enhance initial ADE signals generated from AERS, and to provide a decision support mechanism to facilitate the identification of novel adverse events. RESULTS: The method achieved a significant improvement in precision in identifying known ADE, and a more than twofold signal enhancement when applied to the ADE rhabdomyolysis. The simplicity of the method assists in highlighting the etiology of the ADE by identifying structurally similar drugs. A set of drugs with strong evidence from both AERS and molecular fingerprint-based modeling is constructed for further analysis. CONCLUSION: The results demonstrate that the proposed methodology could be used as a pharmacovigilance decision support tool to facilitate ADE detection.

Although vaccines pose the best means of preventing influenza infection, strain selection and optimal implementation remain difficult due to antigenic drift and a lack of understanding global spread. Detecting viral movement by sequence analysis is complicated by skewed geographic and seasonal distributions in viral isolates. We propose a probabilistic method that accounts for sampling bias through spatiotemporal clustering and modeling regional and seasonal transmission as a binomial process. Analysis of H3N2 not only confirmed East-Southeast Asia as a source of new seasonal variants, but also increased the resolution of observed transmission to a country level. H1N1 data revealed similar viral spread from the tropics. Network analysis suggested China and Hong Kong as the origins of new seasonal H3N2 strains and the United States as a region where increased vaccination would maximally disrupt global spread of the virus. These techniques provide a promising methodology for the analysis of any seasonal virus, as well as for the continued surveillance of influenza.