Artificial intelligence in the life sciences最新文献

Eye irritation and corrosion are fundamental considerations in developing chemicals to be used in or near the eye, from cleaning products to ophthalmic solutions. Unfortunately, animal testing is currently the standard method to identify compounds that cause eye irritation or corrosion. Yet, there is growing pressure on the part of regulatory agencies both in the USA and abroad to develop New Approach Methodologies (NAMs) that help reduce the need for animal testing and address unmet need to modernize safety evaluation of chemical hazards. In furthering the development and applications of computational NAMs in chemical safety assessment, in this study we have collected the largest expertly curated dataset of compounds tested for eye irritation and corrosion, and employed this data to build and validate binary and multi-classification Quantitative Structure-Activity Relationships (QSAR) models that can reliably assess eye irritation/corrosion potential of novel untested compounds. QSAR models were generated with Random Forest (RF) and Multi-Descriptor Read Across (MuDRA) machine learning (ML) methods, and validated using a 5-fold external cross-validation protocol. These models demonstrated high balanced accuracy (CCR of 0.68–0.88), sensitivity (SE of 0.61–0.84), positive predictive value (PPV of 0.65–0.90), specificity (SP of 0.56–0.91), and negative predictive value (NPV of 0.68–0.85). Overall, MuDRA models outperformed RF models and were applied to predict compounds’ irritation/corrosion potential from the Inactive Ingredient Database, which contains components present in FDA-approved drug products, and from the Cosmetic Ingredient Database, the European Commission source of information on cosmetic substances. All models built and validated in this study are publicly available at the STopTox web portal (https://stoptox.mml.unc.edu/). These models can be employed as reliable tools for identifying potential eye irritant/corrosive compounds.

Active learning has been widely used in drug discovery and design in recent years. In this viewpoint, we will briefly summarize applications of AL for drug discovery and propose two potential limitations of research in this field.

Food is considered as a basic need of human being which can be satisfied through farming. Agriculture not only fulfills humans’ basic needs, but also considered as source of employment worldwide. Agriculture is considered as a backbone of economy and source of employment in the developing countries like India. Agriculture contributes 15.4% in the GDP of India. Agriculture activities are broadly categorized into three major areas: pre-harvesting, harvesting and post harvesting. Advancement in area of machine learning has helped improving gains in agriculture. Machine learning is the current technology which is benefiting farmers to minimize the losses in the farming by providing rich recommendations and insights about the crops. This paper presents an extensive survey of latest machine learning application in agriculture to alleviate the problems in the three areas of pre-harvesting, harvesting and post-harvesting. Application of machine learning in agriculture allows more efficient and precise farming with less human manpower with high quality production.

Chemically induced toxicity is the leading cause of recent extinction of honey bees. In this regard, we developed an innovative artificial intelligence-based web app (BeeToxAI) for assessing the acute toxicity of chemicals to Apis mellifera. Initially, we developed and externally validated QSAR models for classification (external set accuracy ∼91%) through the combination of Random Forest and molecular fingerprints to predict the potential for chemicals to cause acute contact toxicity and acute oral toxicity to honey bees. Then, we developed and externally validated regression QSAR models ($R^2$ = 0.75) using Feedforward Neural Networks (FNNs). Afterward, the best models were implemented in the publicly available BeeToxAI web app (http://beetoxai.labmol.com.br/). The outputs of BeeToxAI are: toxicity predictions with estimated confidence, applicability domain estimation, and color-coded maps of relative structure fragment contributions to toxicity. As an additional assessment of BeeToxAI performance, we collected an external set of pesticides with known bee toxicity that were not included in our modeling dataset. BeeToxAI classification models were able to predict four out of five pesticides correctly. The acute contact toxicity model correctly predicted all of the eight pesticides. Here we demonstrate that BeeToxAI can be used as a rapid new approach methodology for predicting acute toxicity of chemicals in honey bees.

Knowing the uncertainty in a prediction is critical when making expensive investment decisions and when patient safety is paramount, but machine learning (ML) models in drug discovery typically only provide a single best estimate and ignore all sources of uncertainty. Predictions from these models may therefore be over-confident, which can put patients at risk and waste resources when compounds that are destined to fail are further developed. Probabilistic predictive models (PPMs) can incorporate all sources of uncertainty and they return a distribution of predicted values that represents the uncertainty in the prediction. We describe seven sources of uncertainty in PPMs: data, distribution function, mean function, variance function, link function(s), parameters, and hyperparameters. We use toxicity prediction as a running example, but the same principles apply for all prediction models. The consequences of ignoring uncertainty and how PPMs account for uncertainty are also described. We aim to make the discussion accessible to a broad non-mathematical audience. Equations are provided to make ideas concrete for mathematical readers (but can be skipped without loss of understanding) and code is available for computational researchers (https://github.com/stanlazic/ML_uncertainty_quantification).

Omics data can be used to identify biological characteristics from genetic to phenotypic levels during the life span of a living being, while molecular interaction networks have a fundamental impact on life activities. Integrating omics data and molecular interaction networks will help researchers delve into comprehensive information hidden in the data. Here, we propose a new multimodal method — AutoGGN — to integrate multi-omics data with molecular interaction networks based on graph convolutional neural networks (GCNs). We evaluated AutoGGN using three classification tasks: single-cell embryonic developmental stage classification, pan-cancer type classification, and breast cancer subtyping. On all three tasks, AutoGGN showed better performance than other methods. This means AutoGGN has the potential to extract insights more effectively by means of integrating molecular interaction networks with multi-omics data. Additionally, in order to provide a better understanding of how our model makes predictions, we utilized the SHAP module and identified the key genes contributing to the classification, providing insight for the design of downstream biological experiments.

Nanopore sequencing from Oxford Nanopore Technologies (ONT) is a promising third-generation sequencing (TGS) technology that generates relatively longer sequencing reads compared to the next-generation sequencing (NGS) technology. A basecaller is a piece of software that translates the original electrical current signals into nucleotide sequences. The accuracy of the basecaller is crucially important to downstream analysis. Bonito is a deep learning-based basecaller recently developed by ONT. Its neural network architecture is composed of a single convolutional layer followed by three stacked bidirectional gated recurrent unit (GRU) layers. Although Bonito has achieved state-of-the-art base calling accuracy, its speed is too slow to be used in production. We therefore developed Fast-Bonito, by using the neural architecture search (NAS) technique to search for a brand-new neural network backbone, and trained it from scratch using several advanced deep learning model training techniques. The new Fast-Bonito model balanced performance in terms of speed and accuracy. Fast-Bonito was 153.8% faster than the original Bonito on NVIDIA V100 GPU. When running on HUAWEI Ascend 910 NPU, Fast-Bonito was 565% faster than the original Bonito. The accuracy of Fast-Bonito was also slightly higher than that of Bonito. We have made Fast-Bonito open source, hoping it will boost the adoption of TGS in both academia and industry.

Advances in image analysis through deep learning have catalyzed the recent use of molecular images in chemoinformatics and drug design for predictive modeling of compound properties and other applications. For image analysis and representation learning from molecular graphs, convolutional neural networks (CNNs) represent a preferred computational architecture. In this work, we have investigated the questions whether functional groups (FGs) and their distinguishing chemical features can be learned from compound images using CNNs of different complexity and whether such knowledge might be transferable to other prediction tasks. We have shown that frequently occurring FGs were comprehensively learned, leading to highly accurate multi-label FG predictions. Furthermore, we have determined that the FG knowledge acquired by CNNs was sufficient for accurate prediction of compound activity cliffs (ACs) via transfer learning. Re-training of FG prediction models on AC data optimized convolutional layer weights and further improved prediction accuracy. Through feature weight analysis and visualization, a rationale was provided for the ability of CNNs to learn FG chemistry and transfer this knowledge for effective AC prediction.

Deep learning has achieved great successes in traditional fields like computer vision (CV), natural language processing (NLP), speech processing, and more. These advancements have greatly inspired researchers in genomics and made deep learning in genomics an exciting and popular topic. The convolutional neural network (CNN) and recurrent neural network (RNN) are frequently used to solve genomic sequencing and prediction problems, and multiple layer perception (MLP) and auto-encoders (AE) are frequently used for genomic profiling data like RNA expression data and gene mutation data. Here, we introduce a new neural network architecture-the residual fully-connected neural network (RFCN)-and describe its advantage in modeling genomic profiling data. We also incorporate AutoML algorithms and implement AutoGenome, an end-to-end, automated deep learning framework for genomic studies. By utilizing the proposed RFCN architecture, automatic hyper-parameter search, and neural architecture search algorithms, AutoGenome can automatically train high-performance deep learning models for various kinds of genomic profiling data. To help researchers better understand the trained models, AutoGenome can assess the importance of different features and export the most critical features for supervised learning tasks and the representative latent vectors for unsupervised learning tasks. We expect AutoGenome will become a popular tool in genomic studies.