We propose the molecular omics network (MOOMIN) a multimodal graph neural network used by AstraZeneca oncologists to predict the synergy of drug combinations for cancer treatment. Our model learns drug representations at multiple scales based on a drug-protein interaction network and metadata. Structural properties of compounds and proteins are encoded to create vertex features for a message-passing scheme that operates on the bipartite interaction graph. Propagated messages form multi-resolution drug representations which we utilized to create drug pair descriptors. By conditioning the drug combination representations on the cancer cell type we define a synergy scoring function that can inductively score unseen pairs of drugs. Experimental results on the synergy scoring task demonstrate that MOOMIN outperforms state-of-the-art graph fingerprinting, proximity preserving node embedding, and existing deep learning approaches. Further results establish that the predictive performance of our model is robust to hyperparameter changes. We demonstrate that the model makes high-quality predictions over a wide range of cancer cell line tissues, out-of-sample predictions can be validated with external synergy databases, and that the proposed model is data efficient at learning.

In this paper, we introduce ChemicalX, a PyTorch-based deep learning library designed for providing a range of state of the art models to solve the drug pair scoring task. The primary objective of the library is to make deep drug pair scoring models accessible to machine learning researchers and practitioners in a streamlined framework. The design of ChemicalX reuses existing high level model training utilities, geometric deep learning, and deep chemistry layers from the PyTorch ecosystem. Our system provides neural network layers, custom pair scoring architectures, data loaders, and batch iterators for end users. We showcase these features with example code snippets and case studies to highlight the characteristics of ChemicalX. A range of experiments on real world drug-drug interaction, polypharmacy side effect, and combination synergy prediction tasks demonstrate that the models available in ChemicalX are effective at solving the pair scoring task. Finally, we show that ChemicalX could be used to train and score machine learning models on large drug pair datasets with hundreds of thousands of compounds on commodity hardware.

Drug discovery is notorious for its low success rates [5]. Despite best research efforts, the majority of drugs fail at early stages of development, even before they enter clinical trials. This phenomenon stems from the inherent complexity of biological systems and our poor understanding of human diseases. To improve that understanding, swaths of data have been generated in recent years. Still, data does not easily translate into knowledge or actionable insights. Here we explore how approaches from the recommendation system domain could help scientists comprehend the ever-growing amount of biomedical facts. The aim of these efforts is to make better drug development decisions, which ultimately result in safe and efficient treatments for patients [3].

Motivation: The secretome denotes the collection of secreted proteins exported outside of the cell. The functional roles of secreted proteins include the maintenance and remodelling of the extracellular matrix as well as signalling between host and non-host cells. These features make secretomes rich reservoirs of biomarkers for disease classification and host-pathogen interaction studies. Common biomarkers are extracellular proteins secreted via classical pathways that can be predicted from sequence by annotating the presence or absence of N-terminal signal peptides. Several heterogeneous command line tools and web-interfaces exist to identify individual motifs, signal sequences and domains that are either characteristic or strictly excluded from secreted proteins. However, a single flexible secretome-prediction workflow that combines all analytic steps is still missing. Results: To bridge this gap the SecretSanta package implements wrapper and parser functions around established command line tools for the integrative prediction of extracellular proteins that are secreted via classical pathways. The modularity of SecretSanta enables users to create tailored pipelines and apply them across the whole tree of life to facilitate comparison of secretomes across multiple species or under various conditions. Availability and implementation: SecretSanta is implemented in the R programming language and is released under GPL-3 license. All functions have been optimized and parallelized to allow large-scale processing of sequences. The open-source code, installation instructions and vignette with use case scenarios can be downloaded from https://github.com/gogleva/SecretSanta. Supplementary information: Supplementary data are available at Bioinformatics online.