Polymorphisms in microRNAs (miRNAs) and their target sites (PolymiRTS) are known to disrupt miRNA function, leading to the development of disease and variation in physiological and behavioral phenotypes. Here, we describe recent updates to the PolymiRTS database (http://compbio.uthsc.edu/miRSNP), an integrated platform for analyzing the functional impact of genetic polymorphisms in miRNA seed regions and miRNA target sites. Recent advances in genomic technologies have made it possible to identify miRNA–mRNA binding sites from direct mapping experiments such as CLASH (cross linking, ligation and sequencing of hybrids). We have integrated data from CLASH experiments in the PolymiRTS database to provide more complete and accurate miRNA–mRNA interactions. Other significant new features include (i) small insertions and deletions in miRNA seed regions and miRNA target sites, (ii) TargetScan context + score differences for assessing the impact of polymorphic miRNA–mRNA interactions and (iii) biological pathways. The browse and search pages of PolymiRTS allow users to explore the relations between the PolymiRTSs and gene expression traits, physiological and behavioral phenotypes, human diseases and biological pathways.

Nonsynonymous single nucleotide polymorphisms (nsSNPs) are prevalent in genomes and are closely associated with inherited diseases. To facilitate identifying disease-associated nsSNPs from a large number of neutral nsSNPs, it is important to develop computational tools to predict the nsSNP's phenotypic effect (disease-associated versus neutral). nsSNPAnalyzer, a web-based software developed for this purpose, extracts structural and evolutionary information from a query nsSNP and uses a machine learning method called Random Forest to predict the nsSNP's phenotypic effect. nsSNPAnalyzer server is available at http://snpanalyzer.utmem.edu/ .

CTCF is a highly conserved transcriptional regulator protein that performs diverse functions such as regulating gene expression and organizing the 3D structure of the genome. Here, we describe recent updates to a database of CTCF-binding sites, CTCFBSDB (http://insulatordb.uthsc.edu/), which now contains almost 15 million CTCF-binding sequences in 10 species. Since the original publication of the database, studies of the 3D structure of the genome, such as those provided by Hi-C experiments, have suggested that CTCF plays an important role in mediating intra- and inter-chromosomal interactions. To reflect this important progress, we have integrated CTCF-binding sites with genomic topological domains defined using Hi-C data. Additionally, the updated database includes new features enabled by new CTCF-binding site data, including binding site occupancy and the ability to visualize overlapping CTCF-binding sites determined in separate experiments.

Polymorphism in microRNA Target Site (PolymiRTS) database is a collection of naturally occurring DNA variations in putative microRNA target sites. PolymiRTSs may affect gene expression and cause variations in complex phenotypes. The database integrates sequence polymorphism, phenotype and expression microarray data, and characterizes PolymiRTSs as potential candidates responsible for the quantitative trait locus (QTL) effects. It is a resource for studying PolymiRTSs and their implications in phenotypic variations. PolymiRTS database can be accessed at Author Webpage.

As artificial intelligence (AI) is increasingly applied to biomedical research and clinical decisions, developing unbiased AI models that work equally well for all ethnic groups is of crucial importance to health disparity prevention and reduction. However, the biomedical data inequality between different ethnic groups is set to generate new health care disparities through data-driven, algorithm-based biomedical research and clinical decisions. Using an extensive set of machine learning experiments on cancer omics data, we find that current prevalent schemes of multiethnic machine learning are prone to generating significant model performance disparities between ethnic groups. We show that these performance disparities are caused by data inequality and data distribution discrepancies between ethnic groups. We also find that transfer learning can improve machine learning model performance for data-disadvantaged ethnic groups, and thus provides an effective approach to reduce health care disparities arising from data inequality among ethnic groups.

The emergence of deep learning has impacted numerous machine learning based applications and research. The reason for its success lies in two main advantages: 1) it provides the ability to learn very complex non-linear relationships between features and 2) it allows one to leverage information from unlabeled data that does not belong to the problem being handled. This paper presents a transfer learning procedure for cancer classification, which uses feature selection and normalization techniques in conjunction with stacked sparse auto-encoders on gene expression data. While classifying any two tumor types, data of other tumor types were used in unsupervised manner to improve the feature representation. The performance of our algorithm was tested on 36 two-class benchmark datasets from the GEMLeR repository. On performing statistical tests, it is clearly ascertained that our algorithm statistically outperforms several generally used cancer classification approaches. The deep learning based molecular disease classification can be used to guide decisions made on the diagnosis and treatment of diseases, and therefore may have important applications in precision medicine.