Search

Knowledge graphs (KGs) represent connections and relationships between real-world entities. We propose a link prediction framework for KGs named Enrichment-Driven GrAph Reasoner (EDGAR), which infers new edges by mining entity-local rules. This approach leverages enrichment analysis, a well-established statistical method used to identify mechanisms common to sets of differentially expressed genes. EDGAR's inference results are inherently explainable and rankable, with p-values indicating the statistical significance of each enrichment-based rule. We demonstrate the framework's effectiveness on a large-scale biomedical KG, ROBOKOP, focusing on drug repurposing for Alzheimer disease (AD) as a case study. Initially, we extracted 14 known drugs from the KG and identified 20 contextual biomarkers through enrichment analysis, revealing functional pathways relevant to shared drug efficacy for AD. Subsequently, using the top 1000 enrichment results, our system identified 1246 additional drug candidates for AD treatment. The top 10 candidates were validated using evidence from medical literature. EDGAR is deployed within ROBOKOP, complete with a web user interface. This is the first study to apply enrichment analysis to large graph completion and drug repurposing., Comment: 10 pages, 5 figures, 4 tables

Expansive Matching of Experts (EMOE) is a novel method that utilizes support-expanding, extrapolatory pseudo-labeling to improve prediction and uncertainty based rejection on out-of-distribution (OOD) points. We propose an expansive data augmentation technique that generates OOD instances in a latent space, and an empirical trial based approach to filter out augmented expansive points for pseudo-labeling. EMOE utilizes a diverse set of multiple base experts as pseudo-labelers on the augmented data to improve OOD performance through a shared MLP with multiple heads (one per expert). We demonstrate that EMOE achieves superior performance compared to state-of-the-art methods on tabular data.

Structure-based virtual screening (SBVS) is a key workflow in computational drug discovery. SBVS models are assessed by measuring the enrichment of known active molecules over decoys in retrospective screens. However, the standard formula for enrichment cannot estimate model performance on very large libraries. Additionally, current screening benchmarks cannot easily be used with machine learning (ML) models due to data leakage. We propose an improved formula for calculating VS enrichment and introduce the BayesBind benchmarking set composed of protein targets that are structurally dissimilar to those in the BigBind training set. We assess current models on this benchmark and find that none perform appreciably better than a KNN baseline., Comment: 10 pages, 4 figures, and 4 tables. The source code is available at https://github.com/molecularmodelinglab/bigbind

Nearest neighbor-based similarity searching is a common task in chemistry, with notable use cases in drug discovery. Yet, some of the most commonly used approaches for this task still leverage a brute-force approach. In practice this can be computationally costly and overly time-consuming, due in part to the sheer size of modern chemical databases. Previous computational advancements for this task have generally relied on improvements to hardware or dataset-specific tricks that lack generalizability. Approaches that leverage lower-complexity searching algorithms remain relatively underexplored. However, many of these algorithms are approximate solutions and/or struggle with typical high-dimensional chemical embeddings. Here we evaluate whether a combination of low-dimensional chemical embeddings and a k-d tree data structure can achieve fast nearest neighbor queries while maintaining performance on standard chemical similarity search benchmarks. We examine different dimensionality reductions of standard chemical embeddings as well as a learned, structurally-aware embedding -- SmallSA -- for this task. With this framework, searches on over one billion chemicals execute in less than a second on a single CPU core, five orders of magnitude faster than the brute-force approach. We also demonstrate that SmallSA achieves competitive performance on chemical similarity benchmarks.

In deep learning for drug discovery, chemical data are often represented as simplified molecular-input line-entry system (SMILES) sequences which allow for straightforward implementation of natural language processing methodologies, one being the sequence-to-sequence autoencoder. However, we observe that training an autoencoder solely on SMILES is insufficient to learn molecular representations that are semantically meaningful, where semantics are defined by the structural (graph-to-graph) similarities between molecules. We demonstrate by example that autoencoders may map structurally similar molecules to distant codes, resulting in an incoherent latent space that does not respect the structural similarities between molecules. To address this shortcoming we propose Semantically-Aware Latent Space Autoencoder (SALSA), a transformer-autoencoder modified with a contrastive task, tailored specifically to learn graph-to-graph similarity between molecules. Formally, the contrastive objective is to map structurally similar molecules (separated by a single graph edit) to nearby codes in the latent space. To accomplish this, we generate a novel dataset comprised of sets of structurally similar molecules and opt for a supervised contrastive loss that is able to incorporate full sets of positive samples. We compare SALSA to its ablated counterparts, and show empirically that the composed training objective (reconstruction and contrastive task) leads to a higher quality latent space that is more 1) structurally-aware, 2) semantically continuous, and 3) property-aware.

Modeling the properties of chemical mixtures is a difficult but important part of any modeling process intended to be applicable to the often messy and impure phenomena of everyday life, including food and environmental safety, healthcare, etc. Part of this difficulty stems from the increased complexity of designing suitable model validation schemes for mixture data, a fact which has been elucidated in previous work only in the case of binary mixture models. We extend these previously defined validation strategies for QSAR modeling of binary mixtures to the more complex case of general, $N$-ary mixtures and argue that these strategies are applicable to many modeling tasks beyond simple chemical mixtures. Additionally, we propose a method of establishing a baseline model performance for each mixture dataset to be in used in model selection comparisons. This baseline is intended to account for the statistical dependence generically present between the properties of mixtures that share constituents. We contend that without such a baseline, estimates of model performance can be dramatically overestimated, and we demonstrate this with multiple case studies using real and simulated data., Comment: 22 pages, 1 figure

Molecular docking aims to predict the 3D pose of a small molecule in a protein binding site. Traditional docking methods predict ligand poses by minimizing a physics-inspired scoring function. Recently, a diffusion model has been proposed that iteratively refines a ligand pose. We combine these two approaches by training a pose scoring function in a diffusion-inspired manner. In our method, PLANTAIN, a neural network is used to develop a very fast pose scoring function. We parameterize a simple scoring function on the fly and use L-BFGS minimization to optimize an initially random ligand pose. Using rigorous benchmarking practices, we demonstrate that our method achieves state-of-the-art performance while running ten times faster than the next-best method. We release PLANTAIN publicly and hope that it improves the utility of virtual screening workflows., Comment: Camera-ready submission to ICML CompBio workshop. 5 pages and 1 figure

Objective: We aim to learn potential novel cures for diseases from unstructured text sources. More specifically, we seek to extract drug-disease pairs of potential cures to diseases by a simple reasoning over the structure of spoken text. Materials and Methods: We use Google Cloud to transcribe podcast episodes of an NPR radio show. We then build a pipeline for systematically pre-processing the text to ensure quality input to the core classification model, which feeds to a series of post-processing steps for obtaining filtered results. Our classification model itself uses a language model pre-trained on PubMed text. The modular nature of our pipeline allows for ease of future developments in this area by substituting higher quality components at each stage of the pipeline. As a validation measure, we use ROBOKOP, an engine over a medical knowledge graph with only validated pathways, as a ground truth source for checking the existence of the proposed pairs. For the proposed pairs not found in ROBOKOP, we provide further verification using Chemotext. Results: We found 30.4% of our proposed pairs in the ROBOKOP database. For example, our model successfully identified that Omeprazole can help treat heartburn.We discuss the significance of this result, showing some examples of the proposed pairs. Discussion and Conclusion: The agreement of our results with the existing knowledge source indicates a step in the right direction. Given the plug-and-play nature of our framework, it is easy to add, remove, or modify parts to improve the model as necessary. We discuss the results showing some examples, and note that this is a potentially new line of research that has further scope to be explored. Although our approach was originally oriented on radio podcast transcripts, it is input-agnostic and could be applied to any source of textual data and to any problem of interest., Comment: initial submission

The traditional paradigm for materials discovery has been recently expanded to incorporate substantial data driven research. With the intent to accelerate the development and the deployment of new technologies, the AFLOW Fleet for computational materials design automates high-throughput first principles calculations, and provides tools for data verification and dissemination for a broad community of users. AFLOW incorporates different computational modules to robustly determine thermodynamic stability, electronic band structures, vibrational dispersions, thermo-mechanical properties and more. The AFLOW data repository is publicly accessible online at aflow.org, with more than 1.7 million materials entries and a panoply of queryable computed properties. Tools to programmatically search and process the data, as well as to perform online machine learning predictions, are also available., Comment: 14 pages, 8 figures

Machine learning approaches, enabled by the emergence of comprehensive databases of materials properties, are becoming a fruitful direction for materials analysis. As a result, a plethora of models have been constructed and trained on existing data to predict properties of new systems. These powerful methods allow researchers to target studies only at interesting materials $\unicode{x2014}$ neglecting the non-synthesizable systems and those without the desired properties $\unicode{x2014}$ thus reducing the amount of resources spent on expensive computations and/or time-consuming experimental synthesis. However, using these predictive models is not always straightforward. Often, they require a panoply of technical expertise, creating barriers for general users. AFLOW-ML (AFLOW $\underline{\mathrm{M}}$achine $\underline{\mathrm{L}}$earning) overcomes the problem by streamlining the use of the machine learning methods developed within the AFLOW consortium. The framework provides an open RESTful API to directly access the continuously updated algorithms, which can be transparently integrated into any workflow to retrieve predictions of electronic, thermal and mechanical properties. These types of interconnected cloud-based applications are envisioned to be capable of further accelerating the adoption of machine learning methods into materials development., Comment: 10 pages, 2 figures

We propose a novel computational strategy for de novo design of molecules with desired properties termed ReLeaSE (Reinforcement Learning for Structural Evolution). Based on deep and reinforcement learning approaches, ReLeaSE integrates two deep neural networks - generative and predictive - that are trained separately but employed jointly to generate novel targeted chemical libraries. ReLeaSE employs simple representation of molecules by their SMILES strings only. Generative models are trained with stack-augmented memory network to produce chemically feasible SMILES strings, and predictive models are derived to forecast the desired properties of the de novo generated compounds. In the first phase of the method, generative and predictive models are trained separately with a supervised learning algorithm. In the second phase, both models are trained jointly with the reinforcement learning approach to bias the generation of new chemical structures towards those with the desired physical and/or biological properties. In the proof-of-concept study, we have employed the ReLeaSE method to design chemical libraries with a bias toward structural complexity or biased toward compounds with either maximal, minimal, or specific range of physical properties such as melting point or hydrophobicity, as well as to develop novel putative inhibitors of JAK2. The approach proposed herein can find a general use for generating targeted chemical libraries of novel compounds optimized for either a single desired property or multiple properties.

Historically, materials discovery has been driven by a laborious trial-and-error process. The growth of materials databases and emerging informatics approaches finally offer the opportunity to transform this practice into data- and knowledge-driven rational design. By using data from the AFLOW repository for high-throughput ab-initio calculations, we have generated Quantitative Materials Structure-Property Relationship (QMSPR) models to predict eight critical electronic and thermomechanical materials properties, such as the metal/insulator classification, band gap energy, bulk and shear moduli, Debye temperature, and heat capacity. The prediction accuracy obtained with these QMSPR models approaches training data for virtually any stoichiometric inorganic crystalline material. The success and universality of these models is attributed to the construction of new materials descriptors---referred to as the universal Property-Labeled Materials Fragments (PLMF). The representation requires only minimal structural input and affords straightforward model interpretation in terms of simple heuristic design rules that guide rational materials design. This study demonstrates the power of materials informatics to dramatically accelerate the search for new materials., Comment: 14 pages, 7 figures