The employment of machine learning (ML) approaches to extract gene expression information from microarray studies has increased in the past years, specially on cancer-related works. However, despite this continuous interest in applying ML in cancer biomedical research, there are no curated repositories focused only on providing quality data sets exclusively for benchmarking and testing of such techniques for cancer research. Thus, in this work, we present the Curated Microarray Database (CuMiDa), a database composed of 78 handpicked microarray data sets for Homo sapiens that were carefully examined from more than 30,000 microarray experiments from the Gene Expression Omnibus using a rigorous filtering criteria. All data sets were individually submitted to background correction, normalization, sample quality analysis and were manually edited to eliminate erroneous probes. All data sets were tested using principal component analysis (PCA) and t-distributed stochastic neighbor embedding (t-SNE) analyses to observe sample division and were additionally tested using various ML approaches to provide a base accuracy for the major techniques employed for microarray data sets. CuMiDa is a database created solely for benchmarking and testing of ML approaches applied to cancer research.

The identification of lead compounds usually includes a step of chemical diversity generation. Its rationale may be supported by both qualitative (SAR) and quantitative (QSAR) approaches, offering models of the putative ligand-receptor interactions. In both scenarios, our understanding of which interactions functional groups can perform is mostly based on their chemical nature (such as electronegativity, volume, melting point, lipophilicity etc.) instead of their dynamics in aqueous, biological solutions (solvent accessibility, lifetime of hydrogen bonds, solvent structure etc.). As a consequence, it is challenging to predict from 2D structures which functional groups will be able to perform interactions with the target receptor, at which intensity and relative abundance in the biological environment, all of which will contribute to ligand potency and intrinsic activity. With this in mind, the aim of this work is to assess properties of aromatic rings, commonly used for drug design, in aqueous solution through molecular dynamics simulations in order to characterize their chemical features and infer their impact in complexation dynamics. For this, common aromatic and heteroaromatic rings were selected and received new atomic charge set based on the direction and module of the dipole moment from MP2/6-31G* calculations, while other topological terms were taken from GROMOS53A6 force field. Afterwards, liquid physicochemical properties were simulated for a calibration set composed by nearly 40 molecules and compared to their respective experimental data, in order to validate each topology. Based on the reliance of the employed strategy, we expanded the dataset to more than 100 aromatic rings. Properties in aqueous solution such as solvent accessible surface area, H-bonds availability, H-bonds residence time, and water structure around heteroatoms were calculated for each ring, creating a database of potential interactions, shedding light on features of drugs in biological solutions, on the structural basis for bioisosterism and on the enthalpic/entropic costs for ligand-receptor complexation dynamics.

The Coronavirus pandemic caused by the novel SARS-CoV-2 has significantly impacted human health and the economy, especially in countries struggling with financial resources for medical testing and treatment, such as Brazil’s case, the third most affected country by the pandemic. In this scenario, machine learning techniques have been heavily employed to analyze different types of medical data, and aid decision making, offering a low-cost alternative. Due to the urgency to fight the pandemic, a massive amount of works are applying machine learning approaches to clinical data, including complete blood count (CBC) tests, which are among the most widely available medical tests. In this work, we review the most employed machine learning classifiers for CBC data, together with popular sampling methods to deal with the class imbalance. Additionally, we describe and critically analyze three publicly available Brazilian COVID-19 CBC datasets and evaluate the performance of eight classifiers and five sampling techniques on the selected datasets. Our work provides a panorama of which classifier and sampling methods provide the best results for different relevant metrics and discuss their impact on future analyses. The metrics and algorithms are introduced in a way to aid newcomers to the field. Finally, the panorama discussed here can significantly benefit the comparison of the results of new ML algorithms.