The Psychiatric Cell Map Initiative: A Convergent Systems Biological Approach to Illuminating Key Molecular Pathways in Neuropsychiatric Disorders.

Although gene discovery in neuropsychiatric disorders, including autism spectrum disorder, intellectual disability, epilepsy, schizophrenia, and Tourette disorder, has accelerated, resulting in a large number of molecular clues, it has proven difficult to generate specific hypotheses without the corresponding datasets at the protein complex and functional pathway level. Here, we describe one path forward-an initiative aimed at mapping the physical and genetic interaction networks of these conditions and then using these maps to connect the genomic data to neurobiology and, ultimately, the clinic. These efforts will include a team of geneticists, structural biologists, neurobiologists, systems biologists, and clinicians, leveraging a wide array of experimental approaches and creating a collaborative infrastructure necessary for long-term investigation. This initiative will ultimately intersect with parallel studies that focus on other diseases, as there is a significant overlap with genes implicated in cancer, infectious disease, and congenital heart defects.

Proximity Graph Networks: Predicting Ligand Affinity with Message Passing Neural Networks.

Message passing neural networks (MPNNs) on molecular graphs generate continuous and differentiable encodings of small molecules with state-of-the-art performance on protein-ligand complex scoring tasks. Here, we describe the proximity graph network (PGN) package, an open-source toolkit that constructs ligand-receptor graphs based on atom proximity and allows users to rapidly apply and evaluate MPNN architectures for a broad range of tasks. We demonstrate the utility of PGN by introducing benchmarks for affinity and docking score prediction tasks. Graph networks generalize better than fingerprint-based models and perform strongly for the docking score prediction task. Overall, MPNNs with proximity graph data structures augment the prediction of ligand-receptor complex properties when ligand-receptor data are available.

Prioritizing Virtual Screening with Interpretable Interaction Fingerprints.

Machine learning-based drug discovery success depends on molecular representation. Yet traditional molecular fingerprints omit both the protein and pointers back to structural information that would enable better model interpretability. Therefore, we propose LUNA, a Python 3 toolkit that calculates and encodes protein-ligand interactions into new hashed fingerprints inspired by Extended Connectivity FingerPrint (ECFP): EIFP (Extended Interaction FingerPrint), FIFP (Functional Interaction FingerPrint), and Hybrid Interaction FingerPrint (HIFP). LUNA also provides visual strategies to make the fingerprints interpretable. We performed three major experiments exploring the fingerprints' use. First, we trained machine learning models to reproduce DOCK3.7 scores using 1 million docked Dopamine D4 complexes. We found that EIFP-4,096 performed (R2 = 0.61) superior to related molecular and interaction fingerprints. Second, we used LUNA to support interpretable machine learning models. Finally, we demonstrate that interaction fingerprints can accurately identify similarities across molecular complexes that other fingerprints overlook. Hence, we envision LUNA and its interface fingerprints as promising methods for machine learning-based virtual screening campaigns. LUNA is freely available at https://github.com/keiserlab/LUNA.

Adding Stochastic Negative Examples into Machine Learning Improves Molecular Bioactivity Prediction.

Multitask deep neural networks learn to predict ligand-target binding by example, yet public pharmacological data sets are sparse, imbalanced, and approximate. We constructed two hold-out benchmarks to approximate temporal and drug-screening test scenarios, whose characteristics differ from a random split of conventional training data sets. We developed a pharmacological data set augmentation procedure, Stochastic Negative Addition (SNA), which randomly assigns untested molecule-target pairs as transient negative examples during training. Under the SNA procedure, drug-screening benchmark performance increases from R2 = 0.1926 ± 0.0186 to 0.4269 ± 0.0272 (122%). This gain was accompanied by a modest decrease in the temporal benchmark (13%). SNA increases in drug-screening performance were consistent for classification and regression tasks and outperformed y-randomized controls. Our results highlight where data and feature uncertainty may be problematic and how leveraging uncertainty into training improves predictions of drug-target relationships.

Zebrafish behavioral profiling identifies multitarget antipsychotic-like compounds.

Many psychiatric drugs act on multiple targets and therefore require screening assays that encompass a wide target space. With sufficiently rich phenotyping and a large sampling of compounds, it should be possible to identify compounds with desired mechanisms of action on the basis of behavioral profiles alone. Although zebrafish (Danio rerio) behavior has been used to rapidly identify neuroactive compounds, it is not clear what types of behavioral assays would be necessary to identify multitarget compounds such as antipsychotics. Here we developed a battery of behavioral assays in larval zebrafish to determine whether behavioral profiles can provide sufficient phenotypic resolution to identify and classify psychiatric drugs. Using the antipsychotic drug haloperidol as a test case, we found that behavioral profiles of haloperidol-treated zebrafish could be used to identify previously uncharacterized compounds with desired antipsychotic-like activities and multitarget mechanisms of action.

Predicted Biological Activity of Purchasable Chemical Space.

Whereas 400 million distinct compounds are now purchasable within the span of a few weeks, the biological activities of most are unknown. To facilitate access to new chemistry for biology, we have combined the Similarity Ensemble Approach (SEA) with the maximum Tanimoto similarity to the nearest bioactive to predict activity for every commercially available molecule in ZINC. This method, which we label SEA+TC, outperforms both SEA and a naïve-Bayesian classifier via predictive performance on a 5-fold cross-validation of ChEMBL's bioactivity data set (version 21). Using this method, predictions for over 40% of compounds (>160 million) have either high significance (pSEA ≥ 40), high similarity (ECFP4MaxTc ≥ 0.4), or both, for one or more of 1382 targets well described by ligands in the literature. Using a further 1347 less-well-described targets, we predict activities for an additional 11 million compounds. To gauge whether these predictions are sensible, we investigate 75 predictions for 50 drugs lacking a binding affinity annotation in ChEMBL. The 535 million predictions for over 171 million compounds at 2629 targets are linked to purchasing information and evidence to support each prediction and are freely available via https://zinc15.docking.org and https://files.docking.org .

Large-scale prediction and testing of drug activity on side-effect targets.

Discovering the unintended 'off-targets' that predict adverse drug reactions is daunting by empirical methods alone. Drugs can act on several protein targets, some of which can be unrelated by conventional molecular metrics, and hundreds of proteins have been implicated in side effects. Here we use a computational strategy to predict the activity of 656 marketed drugs on 73 unintended 'side-effect' targets. Approximately half of the predictions were confirmed, either from proprietary databases unknown to the method or by new experimental assays. Affinities for these new off-targets ranged from 1 nM to 30 µM. To explore relevance, we developed an association metric to prioritize those new off-targets that explained side effects better than any known target of a given drug, creating a drug-target-adverse drug reaction network. Among these new associations was the prediction that the abdominal pain side effect of the synthetic oestrogen chlorotrianisene was mediated through its newly discovered inhibition of the enzyme cyclooxygenase-1. The clinical relevance of this inhibition was borne out in whole human blood platelet aggregation assays. This approach may have wide application to de-risking toxicological liabilities in drug discovery.

In silico molecular comparisons of C. elegans and mammalian pharmacology identify distinct targets that regulate feeding.

Phenotypic screens can identify molecules that are at once penetrant and active on the integrated circuitry of a whole cell or organism. These advantages are offset by the need to identify the targets underlying the phenotypes. Additionally, logistical considerations limit screening for certain physiological and behavioral phenotypes to organisms such as zebrafish and C. elegans. This further raises the challenge of elucidating whether compound-target relationships found in model organisms are preserved in humans. To address these challenges we searched for compounds that affect feeding behavior in C. elegans and sought to identify their molecular mechanisms of action. Here, we applied predictive chemoinformatics to small molecules previously identified in a C. elegans phenotypic screen likely to be enriched for feeding regulatory compounds. Based on the predictions, 16 of these compounds were tested in vitro against 20 mammalian targets. Of these, nine were active, with affinities ranging from 9 nM to 10 µM. Four of these nine compounds were found to alter feeding. We then verified the in vitro findings in vivo through genetic knockdowns, the use of previously characterized compounds with high affinity for the four targets, and chemical genetic epistasis, which is the effect of combined chemical and genetic perturbations on a phenotype relative to that of each perturbation in isolation. Our findings reveal four previously unrecognized pathways that regulate feeding in C. elegans with strong parallels in mammals. Together, our study addresses three inherent challenges in phenotypic screening: the identification of the molecular targets from a phenotypic screen, the confirmation of the in vivo relevance of these targets, and the evolutionary conservation and relevance of these targets to their human orthologs.

Predicting new molecular targets for known drugs.

Although drugs are intended to be selective, at least some bind to several physiological targets, explaining side effects and efficacy. Because many drug-target combinations exist, it would be useful to explore possible interactions computationally. Here we compared 3,665 US Food and Drug Administration (FDA)-approved and investigational drugs against hundreds of targets, defining each target by its ligands. Chemical similarities between drugs and ligand sets predicted thousands of unanticipated associations. Thirty were tested experimentally, including the antagonism of the beta(1) receptor by the transporter inhibitor Prozac, the inhibition of the 5-hydroxytryptamine (5-HT) transporter by the ion channel drug Vadilex, and antagonism of the histamine H(4) receptor by the enzyme inhibitor Rescriptor. Overall, 23 new drug-target associations were confirmed, five of which were potent (<100 nM). The physiological relevance of one, the drug N,N-dimethyltryptamine (DMT) on serotonergic receptors, was confirmed in a knockout mouse. The chemical similarity approach is systematic and comprehensive, and may suggest side-effects and new indications for many drugs.

Prediction and validation of enzyme and transporter off-targets for metformin.

Metformin, an established first-line treatment for patients with type 2 diabetes, has been associated with gastrointestinal (GI) adverse effects that limit its use. Histamine and serotonin have potent effects on the GI tract. The effects of metformin on histamine and serotonin uptake were evaluated in cell lines overexpressing several amine transporters (OCT1, OCT3 and SERT). Metformin inhibited histamine and serotonin uptake by OCT1, OCT3 and SERT in a dose-dependent manner, with OCT1-mediated amine uptake being most potently inhibited (IC50 = 1.5 mM). A chemoinformatics-based method known as Similarity Ensemble Approach predicted diamine oxidase (DAO) as an additional intestinal target of metformin, with an E-value of 7.4 × 10(-5). Inhibition of DAO was experimentally validated using a spectrophotometric assay with putrescine as the substrate. The Ki of metformin for DAO was measured to be 8.6 ± 3.1 mM. In this study, we found that metformin inhibited intestinal amine transporters and DAO at concentrations that may be achieved in the intestine after therapeutic doses. Further studies are warranted to determine the relevance of these interactions to the adverse effects of metformin on the gastrointestinal tract.

Chemical informatics and target identification in a zebrafish phenotypic screen.

Target identification is a core challenge in chemical genetics. Here we use chemical similarity to computationally predict the targets of 586 compounds that were active in a zebrafish behavioral assay. Among 20 predictions tested, 11 compounds had activities ranging from 1 nM to 10,000 nM on the predicted targets. The roles of two of these targets were tested in the original zebrafish phenotype. Prediction of targets from chemotype is rapid and may be generally applicable.

ChromaFactor: deconvolution of single-molecule chromatin organization with non-negative matrix factorization.

The investigation of chromatin organization in single cells holds great promise for identifying causal relationships between genome structure and function. However, analysis of single-molecule data is hampered by extreme yet inherent heterogeneity, making it challenging to determine the contributions of individual chromatin fibers to bulk trends. To address this challenge, we propose ChromaFactor, a novel computational approach based on non-negative matrix factorization that deconvolves single-molecule chromatin organization datasets into their most salient primary components. ChromaFactor provides the ability to identify trends accounting for the maximum variance in the dataset while simultaneously describing the contribution of individual molecules to each component. Applying our approach to two single-molecule imaging datasets across different genomic scales, we find that these primary components demonstrate significant correlation with key functional phenotypes, including active transcription, enhancer-promoter distance, and genomic compartment. ChromaFactor offers a robust tool for understanding the complex interplay between chromatin structure and function on individual DNA molecules, pinpointing which subpopulations drive functional changes and fostering new insights into cellular heterogeneity and its implications for bulk genomic phenomena.

In silico discovery of repetitive elements as key sequence determinants of 3D genome folding.

Natural and experimental genetic variants can modify DNA loops and insulating boundaries to tune transcription, but it is unknown how sequence perturbations affect chromatin organization genome wide. We developed a deep-learning strategy to quantify the effect of any insertion, deletion, or substitution on chromatin contacts and systematically scored millions of synthetic variants. While most genetic manipulations have little impact, regions with CTCF motifs and active transcription are highly sensitive, as expected. Our unbiased screen and subsequent targeted experiments also point to noncoding RNA genes and several families of repetitive elements as CTCF-motif-free DNA sequences with particularly large effects on nearby chromatin interactions, sometimes exceeding the effects of CTCF sites and explaining interactions that lack CTCF. We anticipate that our disruption tracks may be of broad interest and utility as a measure of 3D genome sensitivity, and our computational strategies may serve as a template for biological inquiry with deep learning.

Learning chemical sensitivity reveals mechanisms of cellular response.

Chemical probes interrogate disease mechanisms at the molecular level by linking genetic changes to observable traits. However, comprehensive chemical screens in diverse biological models are impractical. To address this challenge, we developed ChemProbe, a model that predicts cellular sensitivity to hundreds of molecular probes and drugs by learning to combine transcriptomes and chemical structures. Using ChemProbe, we inferred the chemical sensitivity of cancer cell lines and tumor samples and analyzed how the model makes predictions. We retrospectively evaluated drug response predictions for precision breast cancer treatment and prospectively validated chemical sensitivity predictions in new cellular models, including a genetically modified cell line. Our model interpretation analysis identified transcriptome features reflecting compound targets and protein network modules, identifying genes that drive ferroptosis. ChemProbe is an interpretable in silico screening tool that allows researchers to measure cellular response to diverse compounds, facilitating research into molecular mechanisms of chemical sensitivity.

Learning fast and fine-grained detection of amyloid neuropathologies from coarse-grained expert labels.

Precise, scalable, and quantitative evaluation of whole slide images is crucial in neuropathology. We release a deep learning model for rapid object detection and precise information on the identification, locality, and counts of cored plaques and cerebral amyloid angiopathies (CAAs). We trained this object detector using a repurposed image-tile dataset without any human-drawn bounding boxes. We evaluated the detector on a new manually-annotated dataset of whole slide images (WSIs) from three institutions, four staining procedures, and four human experts. The detector matched the cohort of neuropathology experts, achieving 0.64 (model) vs. 0.64 (cohort) average precision (AP) for cored plaques and 0.75 vs. 0.51 AP for CAAs at a 0.5 IOU threshold. It provided count and locality predictions that correlated with gold-standard CERAD-like WSI scoring (p=0.07± 0.10). The openly-available model can quickly score WSIs in minutes without a GPU on a standard workstation.

Learning fast and fine-grained detection of amyloid neuropathologies from coarse-grained expert labels.

Precise, scalable, and quantitative evaluation of whole slide images is crucial in neuropathology. We release a deep learning model for rapid object detection and precise information on the identification, locality, and counts of cored plaques and cerebral amyloid angiopathy (CAA). We trained this object detector using a repurposed image-tile dataset without any human-drawn bounding boxes. We evaluated the detector on a new manually-annotated dataset of whole slide images (WSIs) from three institutions, four staining procedures, and four human experts. The detector matched the cohort of neuropathology experts, achieving 0.64 (model) vs. 0.64 (cohort) average precision (AP) for cored plaques and 0.75 vs. 0.51 AP for CAAs at a 0.5 IOU threshold. It provided count and locality predictions that approximately correlated with gold-standard human CERAD-like WSI scoring (p = 0.07 ± 0.10). The openly-available model can quickly score WSIs in minutes without a GPU on a standard workstation.

Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles.

Machine learning (ML) has increasingly been used to assist and expand current practices in neuropathology. However, generating large imaging datasets with quality labels is challenging in fields which demand high levels of expertise. Further complicating matters is the often seen disagreement between experts in neuropathology-related tasks, both at the case level and at a more granular level. Neurofibrillary tangles (NFTs) are a hallmark pathological feature of Alzheimer disease, and are associated with disease progression which warrants further investigation and granular quantification at a scale not currently accessible in routine human assessment. In this work, we first provide a baseline of annotator/rater agreement for the tasks of Braak NFT staging between experts and NFT detection using both experts and novices in neuropathology. We use a whole-slide-image (WSI) cohort of neuropathology cases from Emory University Hospital immunohistochemically stained for Tau. We develop a workflow for gathering annotations of the early stage formation of NFTs (Pre-NFTs) and mature intracellular (iNFTs) and show ML models can be trained to learn annotator nuances for the task of NFT detection in WSIs. We utilize a model-assisted-labeling approach and demonstrate ML models can be used to aid in labeling large datasets efficiently. We also show these models can be used to extract case-level features, which predict Braak NFT stages comparable to expert human raters, and do so at scale. This study provides a generalizable workflow for various pathology and related fields, and also provides a technique for accomplishing a high-level neuropathology task with limited human annotations.

Deep learning from multiple experts improves identification of amyloid neuropathologies.

Pathologists can label pathologies differently, making it challenging to yield consistent assessments in the absence of one ground truth. To address this problem, we present a deep learning (DL) approach that draws on a cohort of experts, weighs each contribution, and is robust to noisy labels. We collected 100,495 annotations on 20,099 candidate amyloid beta neuropathologies (cerebral amyloid angiopathy (CAA), and cored and diffuse plaques) from three institutions, independently annotated by five experts. DL methods trained on a consensus-of-two strategy yielded 12.6-26% improvements by area under the precision recall curve (AUPRC) when compared to those that learned individualized annotations. This strategy surpassed individual-expert models, even when unfairly assessed on benchmarks favoring them. Moreover, ensembling over individual models was robust to hidden random annotators. In blind prospective tests of 52,555 subsequent expert-annotated images, the models labeled pathologies like their human counterparts (consensus model AUPRC = 0.74 cored; 0.69 CAA). This study demonstrates a means to combine multiple ground truths into a common-ground DL model that yields consistent diagnoses informed by multiple and potentially variable expert opinions.

Trans-channel fluorescence learning improves high-content screening for Alzheimer's disease therapeutics.

In microscopy-based drug screens, fluorescent markers carry critical information on how compounds affect different biological processes. However, practical considerations, such as the labor and preparation formats needed to produce different image channels, hinders the use of certain fluorescent markers. Consequently, completed screens may lack biologically informative but experimentally impractical markers. Here, we present a deep learning method for overcoming these limitations. We accurately generated predicted fluorescent signals from other related markers and validated this new machine learning (ML) method on two biologically distinct datasets. We used the ML method to improve the selection of biologically active compounds for Alzheimer's disease (AD) from a completed high-content high-throughput screen (HCS) that had only contained the original markers. The ML method identified novel compounds that effectively blocked tau aggregation, which had been missed by traditional screening approaches unguided by ML. The method improved triaging efficiency of compound rankings over conventional rankings by raw image channels. We reproduced this ML pipeline on a biologically independent cancer-based dataset, demonstrating its generalizability. The approach is disease-agnostic and applicable across diverse fluorescence microscopy datasets.

Quantifying biogenic bias in screening libraries.

In lead discovery, libraries of 10(6) molecules are screened for biological activity. Given the over 10(60) drug-like molecules thought possible, such screens might never succeed. The fact that they do, even occasionally, implies a biased selection of library molecules. We have developed a method to quantify the bias in screening libraries toward biogenic molecules. With this approach, we consider what is missing from screening libraries and how they can be optimized.