EDeepSSP: Explainable deep neural networks for exact splice sites prediction

Explainable artificial intelligence for omics data: a systematic mapping study

Researchers increasingly turn to explainable artificial intelligence (XAI) to analyze omics data and gain insights into the underlying biological processes. Yet, given the interdisciplinary nature of the field, many findings have only been shared in their respective research community. An overview of XAI for omics data is needed to highlight promising approaches and help detect common issues. Toward this end, we conducted a systematic mapping study. To identify relevant literature, we queried Scopus, PubMed, Web of Science, BioRxiv, MedRxiv and arXiv. Based on keywording, we developed a coding scheme with 10 facets regarding the studies' AI methods, explainability methods and omics data. Our mapping study resulted in 405 included papers published between 2010 and 2023. The inspected papers analyze DNA-based (mostly genomic), transcriptomic, proteomic or metabolomic data by means of neural networks, tree-based methods, statistical methods and further AI methods. The preferred post-hoc explainability methods are feature relevance (n = 166) and visual explanation (n = 52), while papers using interpretable approaches often resort to the use of transparent models (n = 83) or architecture modifications (n = 72). With many research gaps still apparent for XAI for omics data, we deduced eight research directions and discuss their potential for the field. We also provide exemplary research questions for each direction. Many problems with the adoption of XAI for omics data in clinical practice are yet to be resolved. This systematic mapping study outlines extant research on the topic and provides research directions for researchers and practitioners.

A sequence-based two-layer predictor for identifying enhancers and their strength through enhanced feature extraction

Enhancers are short regulatory DNA fragments that are bound with proteins called activators. They are free-bound and distant elements, which play a vital role in controlling gene expression. It is challenging to identify enhancers and their strength due to their dynamic nature. Although some machine learning methods exist to accelerate identification process, their prediction accuracy and efficiency will need more improvement. In this regard, we propose a two-layer prediction model with enhanced feature extraction strategy which does feature combination from improved position-specific amino acid propensity (PSTKNC) method along with Enhanced Nucleic Acid Composition (ENAC) and Composition of k-spaced Nucleic Acid Pairs (CKSNAP). The feature sets from all three feature extraction approaches were concatenated and then sent through a simple artificial neural network (ANN) to accurately identify enhancers in the first layer and their strength in the second layer. Experiments are conducted on benchmark chromatin nine cell lines dataset. A 10-fold cross validation method is employed to evaluate model's performance. The results show that the proposed model gives an outstanding performance with 94.50%, 0.8903 of accuracy and Matthew's correlation coefficient (MCC) in predicting enhancers and fairly does well with independent test also when compared with all other existing methods.

Using amino acids co-occurrence matrices and explainability model to investigate patterns in dengue virus proteins

Background

Dengue is a common vector-borne disease in tropical countries caused by the Dengue virus. This virus may trigger a disease with several symptoms like fever, headache, nausea, vomiting, and muscle pain. Indeed, dengue illness may also present more severe and life-threatening conditions like hemorrhagic fever and dengue shock syndrome. The causes that lead hosts to develop severe infections are multifactorial and not fully understood. However, it is hypothesized that different viral genome signatures may partially contribute to the disease outcome. Therefore, it is plausible to suggest that deeper DENV genetic information analysis may bring new clues about genetic markers linked to severe illness.

Method

Pattern recognition in very long protein sequences is a challenge. To overcome this difficulty, we map protein chains onto matrix data structures that reveal patterns and allow us to classify dengue proteins associated with severe illness outcomes in human hosts. Our analysis uses co-occurrence of amino acids to build the matrices and Random Forests to classify them. We then interpret the classification model using SHAP Values to identify which amino acid co-occurrences increase the likelihood of severe outcomes.

Results

We trained ten binary classifiers, one for each dengue virus protein sequence. We assessed the classifier performance through five metrics: PR-AUC, ROC-AUC, F1-score, Precision and Recall. The highest score on all metrics corresponds to the protein E with a 95% confidence interval. We also compared the means of the classification metrics using the Tukey HSD statistical test. In four of five metrics, protein E was statistically different from proteins M, NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5, showing that E markers has a greater chance to be associated with severe dengue. Furthermore, the amino acid co-occurrence matrix highlight pairs of amino acids within Domain 1 of E protein that may be associated with the classification result.

Conclusion

We show the co-occurrence patterns of amino acids present in the protein sequences that most correlate with severe dengue. This evidence, used by the classification model and verified by statistical tests, mainly associates the E protein with the severe outcome of dengue in human hosts. In addition, we present information suggesting that patterns associated with such severe cases can be found mostly in Domain 1, inside protein E. Altogether, our results may aid in developing new treatments and being the target of debate on new theories regarding the infection caused by dengue in human hosts.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04597-y.

NoAS-DS: Neural optimal architecture search for detection of diverse DNA signals

Neural network architectures are high-performing variable models that can solve many learning tasks. Designing architectures manually require substantial time and also prior knowledge and expertise to develop a high-accuracy model. Most of the architecture search methods are developed over the task of image classification resulting in the building of complex architectures intended for large data inputs such as images. Motivated by the applications of DNA computing in Neural Architecture Search (NAS), we propose NoAS-DS which is specifically built for the architecture search of sequence-based classification tasks. Furthermore, NoAS-DS is applied to the task of predicting binding sites. Unlike other methods that implement only Convolution layers, NoAS-DS, specifically combines Convolution and LSTM layers that helps in the process of automatic architecture building. This hybrid approach helped in achieving high accuracy results on TFBS and RBP datasets which outperformed other models in TF-DNA binding prediction tasks. The best architectures generated by the proposed model can be applied to other DNA datasets of similar nature using transfer learning technique that demonstrates its generalization capability. This greatly reduces the effort required to build new architectures for other prediction tasks.

Predicting novel drug candidates against Covid-19 using generative deep neural networks

The novel Coronavirus outbreak has created a massive economic crisis, and many succumb to death, disturbing the lives of mankind all over the world. Currently, there are no viable treatment for this condition, drug development approaches are being pursued with vigor. The major treatment options are to repurpose existing drugs or to find new ones. Traditional methods for drug discovery take a longer time, so there is an urgent need to develop some alternative techniques that reduces search space for drug candidates. Towards this endeavor, we propose a novel drug discovery method that leverages on long short term memory (LSTM) model to generate novel molecules that are adept at binding with novel Coronavirus protease. Our study demonstrates that the proposed method is able to recreate novel molecules that correlate very much with the properties of trained molecules. Further, we fine-tune the model to generate novel drug-like molecules that are active towards a specific target. We consider 3CLPro, the main protease of novel Coronavirus, as a therapeutic target and demonstrated in silico screening to assess target structural binding affinities with docking simulations. We observed that 80% of generated molecules show docking free energy of less than −5.8 kcal/mol. The top generated drug candidate has the highest binding affinity with a docking score of −8.5 kcal/mol, which is very much lower when compared to approved existing commercial drugs including, Remdesivir. The low binding energy indicates that the generated molecules could be explored as potential drug candidates for Covid-19.