Molecular Descriptors Property Prediction Using Transformer-Based Approach.

In this study, we introduce semi-supervised machine learning models designed to predict molecular properties. Our model employs a two-stage approach, involving pre-training and fine-tuning. Particularly, our model leverages a substantial amount of labeled and unlabeled data consisting of SMILES strings, a text representation system for molecules. During the pre-training stage, our model capitalizes on the Masked Language Model, which is widely used in natural language processing, for learning molecular chemical space representations. During the fine-tuning stage, our model is trained on a smaller labeled dataset to tackle specific downstream tasks, such as classification or regression. Preliminary results indicate that our model demonstrates comparable performance to state-of-the-art models on the chosen downstream tasks from MoleculeNet. Additionally, to reduce the computational overhead, we propose a new approach taking advantage of 3D compound structures for calculating the attention score used in the end-to-end transformer model to predict anti-malaria drug candidates. The results show that using the proposed attention score, our end-to-end model is able to have comparable performance with pre-trained models.

A New Tool to Study the Binding Behavior of Intrinsically Disordered Proteins.

Understanding the binding behavior and conformational dynamics of intrinsically disordered proteins (IDPs) is crucial for unraveling their regulatory roles in biological processes. However, their lack of stable 3D structures poses challenges for analysis. To address this, we propose an algorithm that explores IDP binding behavior with protein complexes by extracting topological and geometric features from the protein surface model. Our algorithm identifies a geometrically favorable binding pose for the IDP and plans a feasible trajectory to evaluate its transition to the docking position. We focus on IDPs from Homo sapiens and Mus-musculus, investigating their interaction with the Plasmodium falciparum (PF) pathogen associated with malaria-related deaths. We compare our algorithm with HawkDock and HDOCK docking tools for quantitative (computation time) and qualitative (binding affinity) measures. Our results indicated that our method outperformed the compared methods in computation performance and binding affinity in experimental conformations.

Strengthening Bioinformatics and Genomics Analyses Skills in Africa for Attainment of the Sustainable Development Goals: Report of the 2nd Conference of the Nigerian Bioinformatics and Genomics Network.

The second conference of the Nigerian Bioinformatics and Genomics Network (NBGN21) was held from October 11 to October 13, 2021. The event was organized by the Nigerian Bioinformatics and Genomics Network. A 1-day genomic analysis workshop on genome-wide association study and polygenic risk score analysis was organized as part of the conference. It was organized primarily as a research capacity building initiative to empower Nigerian researchers to take a leading role in this cutting-edge field of genomic data science. The theme of the conference was "Leveraging Bioinformatics and Genomics for the attainments of the Sustainable Development Goals." The conference used a hybrid approach-virtual and in-person. It served as a platform to bring together 235 registered participants mainly from Nigeria and virtually, from all over the world. NBGN21 had four keynote speakers and four leading Nigerian scientists received awards for their contributions to genomics and bioinformatics development in Nigeria. A total of 100 travel fellowships were awarded to delegates within Nigeria. A major topic of discussion was the application of bioinformatics and genomics in the achievement of the Sustainable Development Goals (SDG3-Good Health and Well-Being, SDG4-Quality Education, and SDG 15-Life on Land [Biodiversity]). In closing, most of the NBGN21 conference participants were interviewed and interestingly they agreed that bioinformatics and genomic analysis of African genomes are vital in identifying population-specific genetic variants that confer susceptibility to different diseases that are endemic in Africa. The knowledge of this can empower African healthcare systems and governments for timely intervention, thereby enhancing good health and well-being.

In-silico analysis reveals druggable single nucleotide polymorphisms in angiotensin 1 converting enzyme involved in the onset of blood pressure.

OBJECTIVE: The Angiotensin 1 converting enzyme (ACE1) gene plays a critical role in regulating blood pressure and thus, it has become a major therapeutic target of antihypertensives. Single nucleotide polymorphisms (SNPs) occurring within a gene most especially at the functional segment of the genes alter the structure-function relationship of that gene. RESULTS: Our study revealed that five nsSNPs of the ACE1 gene were found to be potentially deleterious and damaging and they include rs2229839, rs14507892, rs12709442, and rs4977 at point mutations P351R, R953Q, I1018T, F1051V, and T1187M. The protein stability predictive tools revealed that all the nsSNPs decreased stability of the protein and the Consurf server which estimates the evolutionary conservation profile of a protein showed that three mutants were in the highly conserved region. In conclusion, this study predicted potential druggable deleterious mutants that can be further explored to understand the pathological basis of cardiovascular disease.

Deep Learning Methods in Predicting Gene Expression Levels for the Malaria Parasite.

Malaria is a mosquito-borne disease caused by single-celled blood parasites of the genus Plasmodium. The most severe cases of this disease are caused by the Plasmodium species, Falciparum. Once infected, a human host experiences symptoms of recurrent and intermittent fevers occurring over a time-frame of 48 hours, attributed to the synchronized developmental cycle of the parasite during the blood stage. To understand the regulated periodicity of Plasmodium falciparum transcription, this paper forecast and predict the P. falciparum gene transcription during its blood stage life cycle implementing a well-tuned recurrent neural network with gated recurrent units. Additionally, we also employ a spiking neural network to predict the expression levels of the P. falciparum gene. We provide results of this prediction on multiple genes including potential genes that express possible drug target enzymes. Our results show a high level of accuracy in being able to predict and forecast the expression levels of the different genes.

The Nigerian Bioinformatics and Genomics Network (NBGN): a collaborative platform to advance bioinformatics and genomics in Nigeria.

Africa plays a central importance role in the human origins, and disease susceptibility, agriculture and biodiversity conservation. Nigeria as the most populous and most diverse country in Africa, owing to its 250 ethnic groups and over 500 different native languages is imperative to any global genomic initiative. The newly inaugurated Nigerian Bioinformatics and Genomics Network (NBGN) becomes necessary to facilitate research collaborative activities and foster opportunities for skills' development amongst Nigerian bioinformatics and genomics investigators. NBGN aims to advance and sustain the fields of genomics and bioinformatics in Nigeria by serving as a vehicle to foster collaboration, provision of new opportunities for interactions between various interdisciplinary subfields of genomics, computational biology and bioinformatics as this will provide opportunities for early career researchers. To provide the foundation for sustainable collaborations, the network organises conferences, workshops, trainings and create opportunities for collaborative research studies and internships, recognise excellence, openly share information and create opportunities for more Nigerians to develop the necessary skills to exceed in genomics and bioinformatics. NBGN currently has attracted more than 650 members around the world. Research collaborations between Nigeria, Africa and the West will grow and all stakeholders, including funding partners, African scientists, researchers across the globe, physicians and patients will be the eventual winners. The exponential membership growth and diversity of research interests of NBGN just within weeks of its establishment and the unanticipated attendance of its activities suggest the significant importance of the network to bioinformatics and genomics research in Nigeria.

Adaptive local learning in sampling based motion planning for protein folding.

BACKGROUND: Simulating protein folding motions is an important problem in computational biology. Motion planning algorithms, such as Probabilistic Roadmap Methods, have been successful in modeling the folding landscape. Probabilistic Roadmap Methods and variants contain several phases (i.e., sampling, connection, and path extraction). Most of the time is spent in the connection phase and selecting which variant to employ is a difficult task. Global machine learning has been applied to the connection phase but is inefficient in situations with varying topology, such as those typical of folding landscapes. RESULTS: We develop a local learning algorithm that exploits the past performance of methods within the neighborhood of the current connection attempts as a basis for learning. It is sensitive not only to different types of landscapes but also to differing regions in the landscape itself, removing the need to explicitly partition the landscape. We perform experiments on 23 proteins of varying secondary structure makeup with 52-114 residues. We compare the success rate when using our methods and other methods. We demonstrate a clear need for learning (i.e., only learning methods were able to validate against all available experimental data) and show that local learning is superior to global learning producing, in many cases, significantly higher quality results than the other methods. CONCLUSIONS: We present an algorithm that uses local learning to select appropriate connection methods in the context of roadmap construction for protein folding. Our method removes the burden of deciding which method to use, leverages the strengths of the individual input methods, and it is extendable to include other future connection methods.