Computational Intelligence in Medical Informatics

Sulfur oxidation is one of the oldest known redox processes in our environment mediated by phylogenetically diverse sets of microorganisms. The sulfur oxidation process is mediated mainly by dsr operon which is basically involved in the balancing and utilization of environmental sulfur compounds. DsrMKJOP complex from the dsr operon is the central player of this operon. DsrO is a periplasmic protein which binds FeS clusters responsible for electron transfer to DsrP protein from the dsr operon. DsrP protein is known to be involved in electron transfer to DsrM protein. DsrM protein would then donate the electrons to DsrK protein, the catalytic subunit of this complex. In the present work, we tried to analyze the role of DsrO protein of the dsr operon from the ecologically and industrially important organism Allochromatium vinosum. There are no previous reports that deal with the structural details of the DsrO protein. We predicted the structure of the DsrO protein obtained by homology modeling. The structure of the modeled protein was then docked with various sulfur anion ligands to understand the molecular mechanism of the transportation process of sulfur anion ligands by this DsrMKJOP complex. This study may therefore be considered as a first report of its kind that would therefore enlighten the pathway for analysis of the biochemical mechanism of sulfur oxidation reaction cycle by dsr operon.

Limb girdle muscular dystrophy (OMIM: 613818) is a severe disease in humans, which broadly affects brain development. The disease is caused by T192M mutation in the protein alpha-dystroglycan (α-DG). α-DG is an important component of dystrophin–dystroglycan complex which links extracellular matrices with actin cytoskeleton and thereby maintains signalling cascades essential for the development of tissues and organs. The mutation T192M in α-DG hampers proper glycosylation of α-DG thereby developing limb girdle muscular dystrophy. Prototype virus for Old World Arenaviruses (OWV), Lymphocytic Choriomeningitis virus (LCMV) also uses this α-DG as host cell receptor and invades the host cell causing a disease called Lymphocytic choriomeningitis, an infection to meninges. Thereby, interaction of α-DG and LCMV has become an interesting object of study to predict the mode of the disease onset. In our current work, we have used homology modelling, molecular docking and molecular dynamics (MD) with temperature variation. We have identified significant structural differences between wild type (WT) and mutant (MT) α-DG in terms of spatiotemporal orientations of amino acids. This change in the folding patterns of the WT and MT α-DG has brought forth a different interaction pattern of the WT and MT α-DG with GP1 protein from LCMV as reflected in our docking simulations. Further MD simulations with the complexes over tropical and temperate environment have revealed that MT-α-DG-LCMV GP1 complex is relatively more stable than the wild type counterpart. It has also been found that LCMV GP1 has interacted strongly with mutant α-DG. Our studies therefore has shed light on the structure and molecular interaction pattern of LCMV with MT α-DG and also indicate a possibility of T192M mutant in α-DG making the receptor to interact strongly with LCMV GP1. These insights also provide clues to develop possible therapeutic approaches.

WW domains are the smallest known independently foldable protein structural motifs that are involved in cellular events like protein turnover, splicing, development, and tumor growth control. These motifs bind the polyproline rich ligands. While the WW domains of animal origin are well characterized, the same from plant origin are not well documented yet. Despite the small repertoire of WW proteome of plants (in comparison to animal WW proteome) functional diversity is reported to be equally vivid for plants also. Here, for the first time, we report the structural and functional properties of an Arabidopsis thaliana (At) WW domain containing protein F4JC80 by using homology modeling and docking techniques. Our findings report that the At F4JC80 protein contains two WW domains which bear the standard triple β sheet structure and structurally and functionally resemble Class I WW domains of E3 ubiquitin ligase family but their structural differences impact their polypeptide binding abilities differently.

Multi-parameter patient monitor (MPM) keep track of the condition of a patient in intensive care units (ICU) or general wards using the human vital parameters, heart rate, blood pressure, respiration rate and oxygen saturation (SpO₂). A high accuracy for the overall classification, specificity and sensitivity is extremely important in providing quality health care to the patients. Support vector machine (SVM) is a powerful supervised algorithm that is effectively used in MPMs for classification. A careful study of the vital parameters in a healthy person reveals that there exists an intrinsic relationship between the four vital parameters, for example when heart rate is on the higher side, blood pressure is expected to be on the lower side and vice versa. Hence, it would be highly required to understand the correlation between the vital parameters and to integrate it into the MPM system. In this work, we present the results of the MPM using the SVM as back-end classifier. Further, we use correlation features (feature expansion) along with base parameters in an effort to improve the performance of MPM and note that the performance of the MPM enhanced significantly.

Malaria disease is a major tropical public health problem in the world. The diagnosis of this type of tropical diseases involves several levels of uncertainty and imprecision. It causes severe infection to the brain and prevents brain from its proper functioning. Hence prior detection of the malaria is much essential. Soft Computing Techniques provide excellent methodologies to process the medical data and help medical experts in finding out the nature of illness and to take decision. True data set collection, feature squeezing, and classification are the basic steps followed in designing an expert system. The designed expert system acts with intelligence, prevents erroneous decisions, and produces sharp results in time. This paper discusses on malaria investigation with missing data using rough set rule-based soft computing technique.

Video capsule endoscopy (VCE) has proven to be a pain-free imaging technique of gastrointestinal (GI) tract and provides continuous stream of color imagery. Due to the amount of images captured automatic computer-aided diagnostic (CAD) methods are required to reduce the burden of gastroenterologists. In this work, we propose a fast and efficient method for obtaining segmentations of VCE images automatically without manual supervision. We utilize an efficient active contour without edges model which accounts for topological changes of the mucosal surface when the capsule moves through the GT tract. Comparison with related image segmentation methods indicate we obtain better results in terms of agreement with expert ground-truth boundary markings.

Classification of kinases will provide comparison of related human kinases and insights into kinases functions and evolution. Several algorithms exist for classification and most of them failed to classify when the dimension of feature set large. Selecting the relevant features for classification is significant for variety of reasons like simplification of performance, computational efficiency, and feature interpretability. Generally, feature selection techniques are employed in such cases. However, there has been a limited study on feature selection techniques for classification of biological data. This work tries to determine the impact of feature selection algorithms on classification of kinases. We have used forward greedy feature selection algorithm along with random forest classification algorithm. The performance was evaluated by selecting the feature subset which maximizes Area Under the ROC Curve (AUC). The method identifies the feature subset from the datasets which contains the physiochemical properties of kinases like amino acid, dipeptide, and pseudo amino acid composition. An improvised performance of classification is noted for feature subset than with all the features. Thus, our method indicates that groups of kinases are classifiable with maximum AUC, if good subsets of features are used.

Rheumatoid arthritis (RA) is a long-term systemic inflammatory disease that primarily attacks synovial joints and ultimately leads to their destruction. The disease is characterized by series of processes such as inflammation in the joints, synovial hyperplasia, and cartilage destruction leading to bone erosion. Since RA being a chronic inflammatory complex disease, there is a constant need to develop novel and dynamic treatment to cure the disease. In the present research, network biology and gene expression profiling technology are integrated to predict novel key regulatory molecules, biological pathways, and functional network associated with RA. The microarray datasets of synovial fibroblast (SF) (GSE7669) and macrophages (GSE10500 and GSE8286), which are the primary cells in the synovium and reported as the key players in the pathophysiology of RA, were considered for identification of signature molecules related to RA. The statistical analysis was performed using false discovery rate (FDR), t-test, one-way anova, and Pearson correlation with favorable p-value. The K-core analysis depicted the change in network topology which consisted of up- and downregulated genes network, resulted in six novel meaningful networks with seed genes OAS2, VCAN, CPB1, ZNF516, ACP2, and OLFML2B. Hence, we propose that, differential gene expression network studies will be a standard step to elucidate novel expressed gene(s) globally.