Journal of Bioinformatics and Computational Biology

Thermostability of cellulases can be increased through amino acid substitutions and by protein engineering with predictors of protein thermostability. We have carried out a systematic analysis of the performance of 18 predictors for the engineering of cellulases. The predictors were PoPMuSiC, HoTMuSiC, I-Mutant 2.0, I-Mutant Suite, PremPS, Hotspot, Maestroweb, DynaMut, ENCoM ($\mathrm{\Delta }\mathrm{\Delta }G$ and $\mathrm{\Delta }\mathrm{\Delta }{S}_{\mathrm{\text{vib}}})$, mCSM, SDM, DUET, RosettaDesign, Cupsat (thermal and denaturant approaches), ConSurf, and Voronoia. The highest values of accuracy, F-measure, and MCC were obtained for DynaMut, SDM, RosettaDesign, and PremPS. A combination of the predictors provided an improvement in the performance. F-measure and MCC were improved by 14% and 28%, respectively. Accuracy and sensitivity were also improved by 9% and 20%, respectively, compared to the maximal values of single predictors. The reported values of the performance of the predictors and their combination may aid research in the engineering of thermostable cellulases as well as the further development of thermostability predictors.

Genome rearrangement events are widely used to estimate a minimum-size sequence of mutations capable of transforming a genome into another. The length of this sequence is called distance, and determining it is the main goal in genome rearrangement distance problems. Problems in the genome rearrangement field differ regarding the set of rearrangement events allowed and the genome representation. In this work, we consider the scenario where the genomes share the same set of genes, gene orientation is known or unknown, and intergenic regions (structures between a pair of genes and at the extremities of the genome) are taken into account. We use two models, the first model allows only conservative events (reversals and moves), and the second model includes non-conservative events (insertions and deletions) in the intergenic regions. We show that both models result in NP-hard problems no matter if gene orientation is known or unknown. When the information regarding the orientation of genes is available, we present for both models an approximation algorithm with a factor of 2. For the scenario where this information is unavailable, we propose a 4-approximation algorithm for both models.

Human epidermal growth factor receptor (EGFR) is strongly associated with malignant proliferation and has been established as an attractive therapeutic target of diverse cancers and used as a significant biomarker for tumor diagnosis. Over the past decades, a variety of monoclonal antibodies (mAbs) have been successfully developed to specifically recognize the third subdomain (TSD) of EGFR extracellular domain. Here, the complex crystal structures of EGFR TSD subdomain with its cognate mAbs were examined and compared systematically, revealing a consistent binding mode shared by these mAbs. The recognition site is located on the $\beta$-sheet surface of TSD ladder architecture, from which several hotspot residues that significantly confer both stability and specificity to the recognition were identified, responsible for about half of the total binding potency of mAbs to TSD subdomain. A number of linear peptide mimotopes were rationally designed to mimic these TSD hotspot residues in different orientations and/or in different head-to-tail manners by using an orthogonal threading-through-strand (OTTS) strategy, which, however, are intrinsically disordered in Free State and thus cannot be maintained in a native hotspot-like conformation. A chemical stapling strategy was employed to constrain the free peptides into a double-stranded conformation by introducing a disulfide bond across two strand arms of the peptide mimotopes. Both empirical scoring and fluorescence assay reached an agreement that the stapling can effectively improve the interaction potency of OTTS-designed peptide mimotopes to different mAbs, with binding affinity increase by $>10$-fold. Conformational analysis revealed that the stapled cyclic peptide mimotopes can spontaneously fold into a double-stranded conformation that well threads through all the hotspot residues on TSD $\beta$-sheet surface and exhibits a consistent binding mode with the TSD hotspot site to mAbs.

In this work, we briefly describe our technology developed for computing periodic solutions of time-delay systems and discuss the results of computing periodic solutions for the Marchuk–Petrov model with parameter values, corresponding to hepatitis B infection. We identified the regions in the model parameter space in which an oscillatory dynamics in the form of periodic solutions exists. The respective solutions can be interpreted as active forms of chronic hepatitis B. The period and amplitude of oscillatory solutions were traced along the parameter determining the efficacy of antigen presentation by macrophages for T- and B-lymphocytes in the model.. The oscillatory regimes are characterized by enhanced destruction of hepatocytes as a consequence of immunopathology and temporal reduction of viral load to values which can be a prerequisite of spontaneous recovery observed in chronic HBV infection. Our study presents a first step in a systematic analysis of the chronic HBV infection using Marchuk–Petrov model of antiviral immune response.

Antimicrobial resistance is a major public health concern. Antimicrobial peptides (AMPs) are one of the host defense mechanisms responding efficiently against multidrug-resistant microbes. Since the process of screening AMPs from a large number of peptides is still high-priced and time-consuming, the development of a precise and rapid computer-aided tool is essential for preliminary AMPs selection ahead of laboratory experiments. In this study, we proposed AMPs recognition models using a new peptide encoding method called amino acid index weight (AAIW). Four AMPs recognition models including antimicrobial, antibacterial, antiviral, and antifungal were trained based on datasets combined from the DRAMP and other published databases. These models achieved high performance compared to the preceding AMPs recognition models when evaluated on two independent test sets. All four models yielded over 93% in accuracy and 0.87 in Matthew’s correlation coefficient (MCC). An online AMPs recognition server is accessible at https://amppred-aaiw.com.

Despite an exponential increase in publications on clinical prediction models over recent years, the number of models deployed in clinical practice remains fairly limited. In this paper, we identify common obstacles that impede effective deployment of prediction models in healthcare, and investigate their underlying causes. We observe a key underlying cause behind most obstacles — the improper development and evaluation of prediction models. Inherent heterogeneities in clinical data complicate the development and evaluation of clinical prediction models. Many of these heterogeneities in clinical data are unreported because they are deemed to be irrelevant, or due to privacy concerns. We provide real-life examples where failure to handle heterogeneities in clinical data, or sources of biases, led to the development of erroneous models. The purpose of this paper is to familiarize modeling practitioners with common sources of biases and heterogeneities in clinical data, both of which have to be dealt with to ensure proper development and evaluation of clinical prediction models. Proper model development and evaluation, together with complete and thorough reporting, are important prerequisites for a prediction model to be effectively deployed in healthcare.