Amino acid side chain parameters for correlation studies in biology and pharmacology

Multifunctional host-defense peptides isolated from skin secretions of the banana tree dwelling frog Boana platanera (Hylidae; Hylinae)

Five host-defense peptides (figainin 2PL, hylin PL, raniseptin PL, plasticin PL, and peptide YL) were isolated from norepinephrine-stimulated skin secretions of the banana tree dwelling frog Boana platanera (Hylidae; Hylinae) collected in Trinidad. Raniseptin PL (GVFDTVKKIGKAVGKFALGVAKNYLNS.NH2) and figainin 2PL (FLGTVLKLGKAIAKTVVPMLTNAMQPKQ. NH2) showed potent and rapid bactericidal activity against a range of clinically relevant Gram-positive and Gram-negative ESKAPE + pathogens and Clostridioides difficile. The peptides also showed potent cytotoxic activity (LC50 values < 30 μM) against A549, MDA-MB-231 and HT29 human tumor-derived cell lines but appreciably lower hemolytic activity against mouse erythrocytes (LC50 = 262 ± 14 μM for raniseptin PL and 157 ± 16 μM for figainin 2PL). Hylin PL (FLGLIPALAGAIGNLIK.NH2) showed relatively weak activity against microorganisms but was more hemolytic. The glycine-leucine-rich peptide with structural similarity to the plasticins (GLLSTVGGLVGGLLNNLGL.NH2) and the non-cytotoxic peptide YL (YVPGVIESLL.NH2) lacked antimicrobial and cytotoxic activities. Hylin PL, raniseptinPL and peptide YL stimulated the rate of release of insulin from BRIN-BD11 clonal β-cells at concentrations ≥100 nM. Peptide YL was the most effective (2.3-fold increase compared with basal rate at 1 μM concentration) and may represent a template for the design of a new class of incretin-based anti-diabetic drugs.

Natural solution for the remediation of multi-metal contamination: application of natural amino acids, Pseudomonas fluorescens and Micrococcus yunnanensis to increase the phytoremediation efficiency

Natural amino acids (NAA) have been rarely investigated as chelators, despite their ability to chelate heavy metals (HMs). In the present research, the effects of extracted natural amino acids, as a natural and environmentally friendly chelate agent and the inoculation of Pseudomonas fluorescens (PF) and Micrococcus yunnanensis (MY) bacteria were investigated on some responses of quinoa in a soil polluted with Pb, Ni, Cd, and Zn. Inoculation of PGPR bacteria enhanced plant growth and phytoremediation efficiency. Pb and Cd were higher in quinoa roots, while Ni and Zn were higher in the shoots. The highest efficiencies were observed with NAA treatment and simultaneous inoculation of PF and MY bacteria for Ni, Cd, Pb, and Zn. The highest values of phytoremediation efficiency and uptake efficiency of Ni, Cd, Pb, and Zn were 21.28, 19.11, 14.96 and 18.99 μg g-1, and 31.52, 60.78, 51.89, and 25.33 μg g-1, respectively. Results of present study well demonstrated NAA extracted from blood powder acted as strong chelate agent due to their diversity in size, solubilizing ability, abundant functional groups, and potential in the formation of stable complexes with Ni, Cd, Pb, and Zn, increasing metal availability in soil and improving phytoremediation efficiency in quinoa.

Understanding multicomponent low molecular weight gels from gelators to networks

BACKGROUND

The construction of gels from low molecular weight gelators (LMWG) has been extensively studied in the fields of bio-nanotechnology and other fields. However, the understanding gaps still prevent the prediction of LMWG from the full design of those gel systems. Gels with multicomponent become even more complicated because of the multiple interference effects coexist in the composite gel systems.

AIM OF REVIEW

This review emphasizes systems view on the understanding of multicomponent low molecular weight gels (MLMWGs), and summarizes recent progress on the construction of desired networks of MLMWGs, including self-sorting and co-assembly, as well as the challenges and approaches to understanding MLMWGs, with the hope that the opportunities from natural products and peptides can speed up the understanding process and close the gaps between the design and prediction of structures.

KEY SCIENTIFIC CONCEPTS OF REVIEW

This review is focused on three key concepts. Firstly, understanding the complicated multicomponent gels systems requires a systems perspective on MLMWGs. Secondly, several protocols can be applied to control self-sorting and co-assembly behaviors in those multicomponent gels system, including the certain complementary structures, chirality inducing and dynamic control. Thirdly, the discussion is anchored in challenges and strategies of understanding MLMWGs, and some examples are provided for the understanding of multicomponent gels constructed from small natural products and subtle designed short peptides.

MEnTaT: A machine-learning approach for the identification of mutations to increase protein stability

Enhancing protein thermal stability is important for biomedical and industrial applications as well as in the research laboratory. Here, we describe a simple machine-learning method which identifies amino acid substitutions that contribute to thermal stability based on comparison of the amino acid sequences of homologous proteins derived from bacteria that grow at different temperatures. A key feature of the method is that it compares the sequences based not simply on the amino acid identity, but rather on the structural and physicochemical properties of the side chain. The method accurately identified stabilizing substitutions in three well-studied systems and was validated prospectively by experimentally testing predicted stabilizing substitutions in a polyamine oxidase. In each case, the method outperformed the widely used bioinformatic consensus approach. The method can also provide insight into fundamental aspects of protein structure, for example, by identifying how many sequence positions in a given protein are relevant to temperature adaptation.

From haemadin to haemanorm: Synthesis and characterization of full-length haemadin from the leech Haemadipsa sylvestris and of a novel bivalent, highly potent thrombin inhibitor (haemanorm)

Hirudin from Hirudo medicinalis is a bivalent α-Thrombin (αT) inhibitor, targeting the enzyme active site and exosite-I, and is currently used in anticoagulant therapy along with its simplified analogue hirulog. Haemadin, a small protein (57 amino acids) isolated from the land-living leech Haemadipsa sylvestris, selectively inhibits αT with a potency identical to that of recombinant hirudin (KI  = 0.2 pM), with which it shares a common disulfide topology and overall fold. At variance with hirudin, haemadin targets exosite-II and therefore (besides the free protease) it also blocks thrombomodulin-bound αT without inhibiting the active intermediate meizothrombin, thus offering potential advantages over hirudin. Here, we produced in reasonably high yields and pharmaceutical purity (>98%) wild-type haemadin and the oxidation resistant Met5 → nor-Leucine analogue, both inhibiting αT with a KI of 0.2 pM. Thereafter, we used site-directed mutagenesis, spectroscopic, ligand-displacement, and Hydrogen/Deuterium Exchange-Mass Spectrometry techniques to map the αT regions relevant for the interaction with full-length haemadin and with the synthetic N- and C-terminal peptides Haem(1-10) and Haem(45-57). Haem(1-10) competitively binds to/inhibits αT active site (KI  = 1.9 μM) and its potency was enhanced by 10-fold after Phe3 → β-Naphthylalanine exchange. Conversely to full-length haemadin, haem(45-57) displays intrinsic affinity for exosite-I (KD  = 1.6 μM). Hence, we synthesized a peptide in which the sequences 1-9 and 45-57 were joined together through a 3-Glycine spacer to yield haemanorm, a highly potent (KI  = 0.8 nM) inhibitor targeting αT active site and exosite-I. Haemanorm can be regarded as a novel class of hirulog-like αT inhibitors with potential pharmacological applications.

TROLLOPE: A novel sequence-based stacked approach for the accelerated discovery of linear T-cell epitopes of hepatitis C virus

Hepatitis C virus (HCV) infection is a concerning health issue that causes chronic liver diseases. Despite many successful therapeutic outcomes, no effective HCV vaccines are currently available. Focusing on T cell activity, the primary effector for HCV clearance, T cell epitopes of HCV (TCE-HCV) are considered promising elements to accelerate HCV vaccine efficacy. Thus, accurate and rapid identification of TCE-HCVs is recommended to obtain more efficient therapy for chronic HCV infection. In this study, a novel sequence-based stacked approach, termed TROLLOPE, is proposed to accurately identify TCE-HCVs from sequence information. Specifically, we employed 12 different sequence-based feature descriptors from heterogeneous perspectives, such as physicochemical properties, composition-transition-distribution information and composition information. These descriptors were used in cooperation with 12 popular machine learning (ML) algorithms to create 144 base-classifiers. To maximize the utility of these base-classifiers, we used a feature selection strategy to determine a collection of potential base-classifiers and integrated them to develop the meta-classifier. Comprehensive experiments based on both cross-validation and independent tests demonstrated the superior predictive performance of TROLLOPE compared with conventional ML classifiers, with cross-validation and independent test accuracies of 0.745 and 0.747, respectively. Finally, a user-friendly online web server of TROLLOPE (http://pmlabqsar.pythonanywhere.com/TROLLOPE) has been developed to serve research efforts in the large-scale identification of potential TCE-HCVs for follow-up experimental verification.

Data-mining unveils structure-property-activity correlation of viral infectivity enhancing self-assembling peptides

Gene therapy via retroviral vectors holds great promise for treating a variety of serious diseases. It requires the use of additives to boost infectivity. Amyloid-like peptide nanofibers (PNFs) were shown to efficiently enhance retroviral gene transfer. However, the underlying mode of action of these peptides remains largely unknown. Data-mining is an efficient method to systematically study structure-function relationship and unveil patterns in a database. This data-mining study elucidates the multi-scale structure-property-activity relationship of transduction enhancing peptides for retroviral gene transfer. In contrast to previous reports, we find that not the amyloid fibrils themselves, but rather µm-sized β-sheet rich aggregates enhance infectivity. Specifically, microscopic aggregation of β-sheet rich amyloid structures with a hydrophobic surface pattern and positive surface charge are identified as key material properties. We validate the reliability of the amphiphilic sequence pattern and the general applicability of the key properties by rationally creating new active sequences and identifying short amyloidal peptides from various pathogenic and functional origin. Data-mining-even for small datasets-enables the development of new efficient retroviral transduction enhancers and provides important insights into the diverse bioactivity of the functional material class of amyloids.

Deleting Specific Residues From the HNH Linkers Creates A CRISPR-SpCas9 Variant With High Fidelity and Efficiency

Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR-associated (Cas) systems are immunological defenses used in archaea and bacteria to recognize and destroy DNA from external invaders. The CRISPR-SpCas9 system harnessed from Streptococcus pyogenes (SpCas9) has become the most widely utilized genome editing tool and shows promise for clinical application. However, the off-target effect is still the major challenge for the genome editing of CRISPR-SpCas9. Based on analysis of the structure and cleavage procedures, we proposed two strategies to modify the SpCas9 structure and reduce off-target effects. Shortening the HNH or REC3 linkers (Strategy #1) aimed to move the primary position of HNH or REC3 far away from the single-guide RNA (sgRNA)/DNA hybrid (hybrid), while elongating the helix around the sgRNA (Strategy #2) aimed to strengthen the contacts between SpCas9 and the sgRNA/DNA. We designed 11 SpCas9 variants (variant No.1- variant No.11) and verified their efficiencies on the classic genome site EMX1-1, EMX1-1-OT1, and EMX1-1-OT2. The top three effective SpCas9 variants, variant No.1, variant No.2, and variant No.5, were additionally validated on other genome sites. The further selected variant No.1 was compared with two previous SpCas9 variants, HypaCas9 (a hyper-accurate Cas9 variant released in 2017) and eSpCas9 (1.1) (an "enhanced specificity" SpCas9 variant released in 2016), on two genome sites, EMX1-1 and FANCF-1. The results revealed that the deletion of Thr769 and Gly906 could substantially decrease off-target effects, while maintaining robust on-target efficiency in most of the selected genome sites.

Peptidomic analysis of the host-defense peptides in skin secretions of the Amazon River frog Lithobates palmipes (Ranidae)

Skin secretions of certain frog species represent a source of host-defense peptides (HDPs) with therapeutic potential and their primary structures provide insight into taxonomic and phylogenetic relationships. Peptidomic analysis was used to characterize the HDPs in norepinephrine-stimulated skin secretions from the Amazon River frog Lithobates palmipes (Ranidae) collected in Trinidad. A total of ten peptides were purified and identified on the basis of amino acid similarity as belonging to the ranatuerin-2 family (ranatuerin-2PMa, -2PMb, -2PMc, and-2PMd), the brevinin-1 family (brevinin-1PMa, -1PMb, -1PMc and des(8-14)brevinin-1PMa) and the temporin family (temporin-PMa in C-terminally amidated and non-amidated forms). Deletion of the sequence VAAKVLP from brevinin-1PMa (FLPLIAGVAAKVLPKIFCAISKKC) in des[(8-14)brevinin-1PMa resulted in a 10-fold decrease in potency against Staphylococcus aureus (MIC = 31 μM compared with 3 μM) and a > 50-fold decrease in hemolytic activity but potency against Echerichia coli was maintained (MIC = 62.5 μM compared with 50 μM). Temporin-PMa (FLPFLGKLLSGIF.NH₂) inhibited growth of S. aureus (MIC = 16 μM) but the non-amidated form of the peptide lacked antimicrobial activity. Cladistic analysis based upon the primary structures of ranaturerin-2 peptides supports the division of New World frogs of the family Ranidae into the genera Lithobates and Rana. A sister-group relationship between L. palmipes and Warszewitsch's frog Lithobates warszewitschii is indicated within a clade that includes the Tarahumara frog Lithobates tarahumarae. The study has provided further evidence that peptidomic analysis of HDPs in frog skin secretions is a valuable approach to elucidation of the evolutionary history of species within a particular genus.

dSCOPE: a software to detect sequences critical for liquid-liquid phase separation

Membrane-based cells are the fundamental structural and functional units of organisms, while evidences demonstrate that liquid-liquid phase separation (LLPS) is associated with the formation of membraneless organelles, such as P-bodies, nucleoli and stress granules. Many studies have been undertaken to explore the functions of protein phase separation (PS), but these studies lacked an effective tool to identify the sequence segments that critical for LLPS. In this study, we presented a novel software called dSCOPE (http://dscope.omicsbio.info) to predict the PS-driving regions. To develop the predictor, we curated experimentally identified sequence segments that can drive LLPS from published literature. Then sliding sequence window based physiological, biochemical, structural and coding features were integrated by random forest algorithm to perform prediction. Through rigorous evaluation, dSCOPE was demonstrated to achieve satisfactory performance. Furthermore, large-scale analysis of human proteome based on dSCOPE showed that the predicted PS-driving regions enriched various protein post-translational modifications and cancer mutations, and the proteins which contain predicted PS-driving regions enriched critical cellular signaling pathways. Taken together, dSCOPE precisely predicted the protein sequence segments critical for LLPS, with various helpful information visualized in the webserver to facilitate LLPS-related research.

Expanding the phenotypic spectrum of KCNK4: From syndromic neurodevelopmental disorder to rolandic epilepsy

The KCNK4 gene, predominantly distributed in neurons, plays an essential role in controlling the resting membrane potential and regulating cellular excitability. Previously, only two variants were identified to be associated with human disease, facial dysmorphism, hypertrichosis, epilepsy, intellectual/developmental delay, and gingival overgrowth (FHEIG) syndrome. In this study, we performed trio-based whole exon sequencing (WES) in a cohort of patients with epilepsy. Two de novo likely pathogenic variants were identified in two unrelated cases with heterogeneous phenotypes, including one with Rolandic epilepsy and one with the FHEIG syndrome. The two variants were predicted to be damaged by the majority of in silico algorithms. These variants showed no allele frequencies in controls and presented statistically higher frequencies in the case cohort than that in controls. The FHEIG syndrome-related variants were all located in the region with vital functions in stabilizing the conductive conformation, while the Rolandic epilepsy-related variant was distributed in the area with less impact on the conductive conformation. This study expanded the genetic and phenotypic spectrum of KCNK4. Phenotypic variations of KCNK4 are potentially associated with the molecular sub-regional effects. Carbamazepine/oxcarbazepine and valproate may be effective antiepileptic drugs for patients with KCNK4 variants.

Variation in the ACE2 receptor has limited utility for SARS-CoV-2 host prediction

Transmission of SARS-CoV-2 from humans to other species threatens wildlife conservation and may create novel sources of viral diversity for future zoonotic transmission. A variety of computational heuristics have been developed to pre-emptively identify susceptible host species based on variation in the angiotensin-converting enzyme 2 (ACE2) receptor used for viral entry. However, the predictive performance of these heuristics remains unknown. Using a newly compiled database of 96 species, we show that, while variation in ACE2 can be used by machine learning models to accurately predict animal susceptibility to sarbecoviruses (accuracy = 80.2%, binomial confidence interval [CI]: 70.8–87.6%), the sites informing predictions have no known involvement in virus binding and instead recapitulate host phylogeny. Models trained on host phylogeny alone performed equally well (accuracy = 84.4%, CI: 75.5–91.0%) and at a level equivalent to retrospective assessments of accuracy for previously published models. These results suggest that the predictive power of ACE2-based models derives from strong correlations with host phylogeny rather than processes which can be mechanistically linked to infection biology. Further, biased availability of ACE2 sequences misleads projections of the number and geographic distribution of at-risk species. Models based on host phylogeny reduce this bias, but identify a very large number of susceptible species, implying that model predictions must be combined with local knowledge of exposure risk to practically guide surveillance. Identifying barriers to viral infection or onward transmission beyond receptor binding and incorporating data which are independent of host phylogeny will be necessary to manage the ongoing risk of establishment of novel animal reservoirs of SARS-CoV-2.

Protein-protein interaction prediction with deep learning: A comprehensive review

Most proteins perform their biological function by interacting with themselves or other molecules. Thus, one may obtain biological insights into protein functions, disease prevalence, and therapy development by identifying protein-protein interactions (PPI). However, finding the interacting and non-interacting protein pairs through experimental approaches is labour-intensive and time-consuming, owing to the variety of proteins. Hence, protein-protein interaction and protein-ligand binding problems have drawn attention in the fields of bioinformatics and computer-aided drug discovery. Deep learning methods paved the way for scientists to predict the 3-D structure of proteins from genomes, predict the functions and attributes of a protein, and modify and design new proteins to provide desired functions. This review focuses on recent deep learning methods applied to problems including predicting protein functions, protein-protein interaction and their sites, protein-ligand binding, and protein design.

Ribosome selectivity and nascent chain context in modulating the incorporation of fluorescent non-canonical amino acid into proteins

Fluorescence reporter groups are important tools to study the structure and dynamics of proteins. Genetic code reprogramming allows for cotranslational incorporation of non-canonical amino acids at any desired position. However, cotranslational incorporation of bulky fluorescence reporter groups is technically challenging and usually inefficient. Here we analyze the bottlenecks for the cotranslational incorporation of NBD-, BodipyFL- and Atto520-labeled Cys-tRNA^Cys into a model protein using a reconstituted in-vitro translation system. We show that the modified Cys-tRNA^Cys can be rejected during decoding due to the reduced ribosome selectivity for the modified aa-tRNA and the competition with native near-cognate aminoacyl-tRNAs. Accommodation of the modified Cys-tRNA^Cys in the A site of the ribosome is also impaired, but can be rescued by one or several Gly residues at the positions −1 to −4 upstream of the incorporation site. The incorporation yield depends on the steric properties of the downstream residue and decreases with the distance from the protein N-terminus to the incorporation site. In addition to the full-length translation product, we find protein fragments corresponding to the truncated N-terminal peptide and the C-terminal fragment starting with a fluorescence-labeled Cys arising from a StopGo-like event due to a defect in peptide bond formation. The results are important for understanding the reasons for inefficient cotranslational protein labeling with bulky reporter groups and for designing new approaches to improve the yield of fluorescence-labeled protein.

On the Prediction of In Vitro Arginine Glycation of Short Peptides Using Artificial Neural Networks

One of the hallmarks of diabetes is an increased modification of cellular proteins. The most prominent type of modification stems from the reaction of methylglyoxal with arginine and lysine residues, leading to structural and functional impairments of target proteins. For lysine glycation, several algorithms allow a prediction of occurrence; thus, making it possible to pinpoint likely targets. However, according to our knowledge, no approaches have been published for predicting the likelihood of arginine glycation. There are indications that arginine and not lysine is the most prominent target for the toxic dialdehyde. One of the reasons why there is no arginine glycation predictor is the limited availability of quantitative data. Here, we used a recently published high-quality dataset of arginine modification probabilities to employ an artificial neural network strategy. Despite the limited data availability, our results achieve an accuracy of about 75% of correctly predicting the exact value of the glycation probability of an arginine-containing peptide without setting thresholds upon whether it is decided if a given arginine is modified or not. This contribution suggests a solution for predicting arginine glycation of short peptides.

Flexible customization of the self-assembling abilities of short elastin-like peptide Fn analogs by substituting N-terminal amino acids

Elastin-like peptides (ELPs) are thermoresponsive biopolymers inspired by the characteristic repetitive sequences of natural elastin. As ELPs exhibit temperature-dependent reversible self-assembly, they are expected to be biocompatible thermoresponsive materials for drug delivery carriers. One of the most widely studied ELPs in this field is the repetitive pentapeptide, (VPGXG)_n . We previously reported that phenylalanine-containing ELP (Fn) analogs, in which the former Val residue of the repetitive sequence (VPGVG)_n is replaced by Phe, show coacervation with a short chain length (n = 5). Owing to their short sequences, Fn analogs are easily modified in amino acid sequences via simple chemical synthesis, and are useful for investigating the relationship between peptide sequences and temperature responsiveness. In this study, we developed Fn analogs by replacing Phe residue(s) with other amino acids or introducing another amino acid at the N-terminus. The temperature responsiveness of the Fn analogs changed drastically with the substitution of a single Phe residue, suggesting that aromatic amino acids play an important role in their self-assembly. In addition, the self-assembling ability of Fn was enhanced by increasing the bulkiness of the N-terminal amino acids. Therefore, the N-terminal residue was considered to be important for hydrophobicity-induced intermolecular interactions between the peptides during coacervation.

Comprehensive Assessment of Indian Variations in the Druggable Kinome Landscape Highlights Distinct Insights at the Sequence, Structure and Pharmacogenomic Stratum

India confines more than 17% of the world’s population and has a diverse genetic makeup with several clinically relevant rare mutations belonging to many sub-group which are undervalued in global sequencing datasets like the 1000 Genome data (1KG) containing limited samples for Indian ethnicity. Such databases are critical for the pharmaceutical and drug development industry where diversity plays a crucial role in identifying genetic disposition towards adverse drug reactions. A qualitative and comparative sequence and structural study utilizing variant information present in the recently published, largest curated Indian genome database (IndiGen) and the 1000 Genome data was performed for variants belonging to the kinase coding genes, the second most targeted group of drug targets. The sequence-level analysis identified similarities and differences among different populations based on the nsSNVs and amino acid exchange frequencies whereas a comparative structural analysis of IndiGen variants was performed with pathogenic variants reported in UniProtKB Humsavar data. The influence of these variations on structural features of the protein, such as structural stability, solvent accessibility, hydrophobicity, and the hydrogen-bond network was investigated. In-silico screening of the known drugs to these Indian variation-containing proteins reveals critical differences imparted in the strength of binding due to the variations present in the Indian population. In conclusion, this study constitutes a comprehensive investigation into the understanding of common variations present in the second largest population in the world and investigating its implications in the sequence, structural and pharmacogenomic landscape. The preliminary investigation reported in this paper, supporting the screening and detection of ADRs specific to the Indian population could aid in the development of techniques for pre-clinical and post-market screening of drug-related adverse events in the Indian population.

SNARER: new molecular descriptors for SNARE proteins classification

Background

SNARE proteins play an important role in different biological functions. This study aims to investigate the contribution of a new class of molecular descriptors (called SNARER) related to the chemical-physical properties of proteins in order to evaluate the performance of binary classifiers for SNARE proteins.

Results

We constructed a SNARE proteins balanced dataset, D128, and an unbalanced one, DUNI, on which we tested and compared the performance of the new descriptors presented here in combination with the feature sets (GAAC, CTDT, CKSAAP and 188D) already present in the literature. The machine learning algorithms used were Random Forest, k-Nearest Neighbors and AdaBoost and oversampling and subsampling techniques were applied to the unbalanced dataset. The addition of the SNARER descriptors increases the precision for all considered ML algorithms. In particular, on the unbalanced DUNI dataset the accuracy increases in parallel with the increase in sensitivity while on the balanced dataset D128 the accuracy increases compared to the counterpart without the addition of SNARER descriptors, with a strong improvement in specificity. Our best result is the combination of our descriptors SNARER with CKSAAP feature on the dataset D128 with 92.3% of accuracy, 90.1% for sensitivity and 95% for specificity with the RF algorithm.

Conclusions

The performed analysis has shown how the introduction of molecular descriptors linked to the chemical-physical and structural characteristics of the proteins can improve the classification performance. Additionally, it was pointed out that performance can change based on using a balanced or unbalanced dataset. The balanced nature of training can significantly improve forecast accuracy.

Predicting the functional impact of KCNQ1 variants with artificial neural networks

Recent advances in experimental and computational protein structure determination have provided access to high-quality structures for most human proteins and mutants thereof. However, linking changes in structure in protein mutants to functional impact remains an active area of method development. If successful, such methods can ultimately assist physicians in taking appropriate treatment decisions. This work presents three artificial neural network (ANN)-based predictive models that classify four key functional parameters of KCNQ1 variants as normal or dysfunctional using PSSM-based evolutionary and/or biophysical descriptors. Recent advances in predicting protein structure and variant properties with artificial intelligence (AI) rely heavily on the availability of evolutionary features and thus fail to directly assess the biophysical underpinnings of a change in structure and/or function. The central goal of this work was to develop an ANN model based on structure and physiochemical properties of KCNQ1 potassium channels that performs comparably or better than algorithms using only on PSSM-based evolutionary features. These biophysical features highlight the structure-function relationships that govern protein stability, function, and regulation. The input sensitivity algorithm incorporates the roles of hydrophobicity, polarizability, and functional densities on key functional parameters of the KCNQ1 channel. Inclusion of the biophysical features outperforms exclusive use of PSSM-based evolutionary features in predicting activation voltage dependence and deactivation time. As AI is increasingly applied to problems in biology, biophysical understanding will be critical with respect to 'explainable AI', i.e., understanding the relation of sequence, structure, and function of proteins. Our model is available at www.kcnq1predict.org.

Lightweight ProteinUnet2 network for protein secondary structure prediction: a step towards proper evaluation

Background

The prediction of protein secondary structures is a crucial and significant step for ab initio tertiary structure prediction which delivers the information about proteins activity and functions. As the experimental methods are expensive and sometimes impossible, many SS predictors, mainly based on different machine learning methods have been proposed for many years. Currently, most of the top methods use evolutionary-based input features produced by PSSM and HHblits software, although quite recently the embeddings—the new description of protein sequences generated by language models (LM) have appeared that could be leveraged as input features. Apart from input features calculation, the top models usually need extensive computational resources for training and prediction and are barely possible to run on a regular PC. SS prediction as the imbalanced classification problem should not be judged by the commonly used Q3/Q8 metrics. Moreover, as the benchmark datasets are not random samples, the classical statistical null hypothesis testing based on the Neyman–Pearson approach is not appropriate.

Results

We present a lightweight deep network ProteinUnet2 for SS prediction which is based on U-Net convolutional architecture and evolutionary-based input features (from PSSM and HHblits) as well as SPOT-Contact features. Through an extensive evaluation study, we report the performance of ProteinUnet2 in comparison with top SS prediction methods based on evolutionary information (SAINT and SPOT-1D). We also propose a new statistical methodology for prediction performance assessment based on the significance from Fisher–Pitman permutation tests accompanied by practical significance measured by Cohen’s effect size.

Conclusions

Our results suggest that ProteinUnet2 architecture has much shorter training and inference times while maintaining results similar to SAINT and SPOT-1D predictors. Taking into account the relatively long times of calculating evolutionary-based features (from PSSM in particular), it would be worth conducting the predictive ability tests on embeddings as input features in the future. We strongly believe that our proposed here statistical methodology for the evaluation of SS prediction results will be adopted and used (and even expanded) by the research community.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04623-z.

Tracking the amino acid changes of spike proteins across diverse host species of severe acute respiratory syndrome coronavirus 2

Knowledge of the host-specific properties of the spike protein is of crucial importance to understand the adaptability of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) to infect multiple species and alter transmissibility, particularly in humans. Here, we propose a spike protein predictor SPIKES incorporating with an inheritable bi-objective combinatorial genetic algorithm to identify the biochemical properties of spike proteins and determine their specificity to human hosts. SPIKES identified 20 informative physicochemical properties of the spike protein, including information measures for alpha helix and relative mutability, and amino acid and dipeptide compositions, which have shown compositional difference at the amino acid sequence level between human and diverse animal coronaviruses. We suggest that alterations of these amino acids between human and animal coronaviruses may provide insights into the development and transmission of SARS-CoV-2 in human and other species and support the discovery of targeted antiviral therapies.

•

Differences exist in the amino acids within the S protein of diverse host species CoVs

•

We developed SPIKES to identify informative properties of S protein

•

SARS-CoV-2 variants have amino acid changes that alter infection and transmission

•

The SPIKES identified changes in S protein properties from animal to human host CoVs

Optimizing CSP1 Analogs for Modulating Quorum Sensing in Streptococcus pneumoniae with Bulky, Hydrophobic Nonproteogenic Amino Acid Substitutions

The prompt appearance of multiantibiotic-resistant bacteria necessitates finding alternative treatments that can attenuate bacterial infections while minimizing the rate of antibiotic resistance development. Streptococcus pneumoniae, a notorious human pathogen, is responsible for severe antibiotic-resistant infections. Its pathogenicity is influenced by a cell-density communication system, termed quorum sensing (QS). As a result, controlling QS through the development of peptide-based QS modulators may serve to attenuate pneumococcal infections. Herein, we set out to evaluate the impact of the introduction of bulkier, nonproteogenic side-chain residues on the hydrophobic binding face of CSP1 to optimize receptor-binding interactions in both of the S. pneumoniae specificity groups. Our results indicate that these substitutions optimize the peptide–protein binding interactions, yielding several pneumococcal QS modulators with high potency. Moreover, pharmacological evaluation of lead analogs revealed that the incorporation of nonproteogenic amino acids increased the peptides’ half-life towards enzymatic degradation while remaining nontoxic. Overall, our data convey key considerations for SAR using nonproteogenic amino acids, which provide analogs with better pharmacological properties.

The prompt appearance of multiantibiotic-resistant bacteria necessitates finding alternative treatments that can attenuate bacterial infections while minimizing the rate of antibiotic resistance development.

Metal-Peptide Nonafoil Knots and Decafoil Supercoils

Despite the frequent occurrence of knotted frameworks in protein structures, the latent potential of peptide strands to form entangled structures is rarely discussed in peptide chemistry. Here we report the construction of highly entangled molecular topologies from Ag(I) ions and tripeptide ligands. The efficient entanglement of metal-peptide strands and the wide scope for design of the amino acid side chains in these ligands enabled the construction of metal-peptide 9₁ torus knots and 10₁² torus links. Moreover, steric control of the peptide side chain induced ring opening and twisting of the torus framework, which resulted in an infinite toroidal supercoil nanostructure.

Mechanistic process understanding of the self-assembling behaviour of asymmetric bolaamphiphilic short-peptides and their templating for silica and titania nanomaterials

Investigation of the self-assembly of peptides is critically important to clarify certain biophysical phenomena, fulfill some biological functions, and construct functional materials. However, it is still a challenge to precisely predict the self-assembled structures of peptides because of their complicated driving forces and various assembling pathways. In this work, to elucidate the effects of noncovalent interactions including hydrogen bonding, molecular geometry, and hydrophobic and electrostatic interactions on the peptide self-assembly, a series of asymmetric bolaamphiphilic short peptides consisting of Ac-EI₃K-NH₂ (EI₃K), Ac-EI₄K-NH₂ (EI₄K), Ac-KI₃E-NH₂ (KI₃E) and Ac-KI₄E-NH₂ (KI₄E) were designed and their self-assembling behaviors at different solution pH values were investigated systematically. The peptides self-assembled into twisted nanofibers under most conditions except for EI₄K in a strongly alkaline solution and KI₄E under a strongly acidic condition, in which they self-assembled into nanotubes via helical monolayer nanosheet intermediates. In particular, KI₄E nanotubes are formed under acidic conditions, and its diameters are ∼500 nm much greater than most of the self-assembled structures from bolaamphiphilic peptides. Moreover, reversible morphological transition between the nanotubes and twisted nanofibers was observed with the change in solution pH. Such tunable self-assembled structures and switchable surface properties of the asymmetric bolaamphiphilic short-peptides allow them to be used as templates to construct advanced materials. Silica and titania nanomaterials faithful to the peptide templates in morphology were prepared at ambient temperature. This work clearly elucidates the effects of noncovalent interactions on the peptide self-assembly and also provides new insights into the design and preparation of complicated inorganic materials from tunable organic templates.

Machine Learning Methods in Prediction of Protein Palmitoylation Sites: A Brief Review

Protein palmitoylation is a fundamental and reversible post-translational lipid modification that involves a series of biological processes. Although a large number of experimental studies have explored the molecular mechanism behind the palmitoylation process, the computational methods has attracted much attention for its good performance in predicting palmitoylation sites compared with expensive and time-consuming biochemical experiments. The prediction of protein palmitoylation sites is helpful to reveal its biological mechanism. Therefore, the research on the application of machine learning methods to predict palmitoylation sites has become a hot topic in bioinformatics and promoted the development in the related fields. In this review, we briefly introduced the recent development in predicting protein palmitoylation sites by using machine learningbased methods and discussed their benefits and drawbacks. The perspective of machine learning-based methods in predicting palmitoylation sites was also provided. We hope the review could provide a guide in related fields.

UMAP-DBP: An Improved DNA-Binding Proteins Prediction Method Based on Uniform Manifold Approximation and Projection

DNA-binding proteins play a vital role in cellular processes. It is an extremely urgent to develop a high-throughput method for efficiently identifying DNA-binding proteins. According to the current research situation, some methods in machine learning and deep learning show excellent computational speed and accuracy, which are worthy of application. In this work, a novel predictor was proposed to predict DNA binding proteins called UMAP-DBP. Firstly, the feature extraction of primary protein sequence was realized based on physicochemical distance transformation, Profile-based auto-cross covariance and General series correlation pseudo amino acid composition. Secondly, uniform manifold approximation and projection (UMAP) and feature importance score methods were used for feature selection; there is a progressive relationship between them. Finally, the Adaboost operation engine with jackknife test were adopted for predicting DNA-binding proteins. For the jackknife test on the BP1075 and BP594, we obtained an overall accuracy of 82.97% and 82.14%, Cohen's kappa (CK) of 0.66 and 0.64, respectively. The results illustrate that a feasible method has been developed for predicting DNA-binding proteins by UMAP and Adaboost. This is the first study in which UMAP has been successfully applied to identify DNA-binding proteins. All the datasets and codes are accessible at https://github.com/Wang-Jinyue/UMAP-DBP .

Capacity for increased surface area in the hydrophobic core of β-sheet peptide bilayer nanoribbons

Amphipathic peptides with amino acids arranged in alternating patterns of hydrophobic and hydrophilic residues efficiently self‐assemble into β‐sheet bilayer nanoribbons. Hydrophobic side chain functionality is effectively buried in the interior of the putative bilayer of these nanoribbons. This study investigates consequences on self‐assembly of increasing the surface area of aromatic side chain groups that reside in the hydrophobic core of nanoribbons derived from Ac‐(XKXE)₂‐NH₂ peptides (X = hydrophobic residue). A series of Ac‐(XKXE)₂‐NH₂ peptides incorporating aromatic amino acids of increasing molecular volume and steric profile (X = phenylalanine [Phe], homophenylalanine [Hph], tryptophan [Trp], 1‐naphthylalanine [1‐Nal], 2‐naphthylalanine [2‐Nal], or biphenylalanine [Bip]) were assessed to determine substitution effects on self‐assembly propensity and on morphology of the resulting nanoribbon structures. Additional studies were conducted to determine the effects of incorporating amino acids of differing steric profile in the hydrophobic core (Ac‐X₁KFEFKFE‐NH₂ and Ac‐(X_1,5KFE)‐NH₂ peptides, X = Trp or Bip). Spectroscopic analysis by circular dichroism (CD) and Fourier transform infrared (FT‐IR) spectroscopy indicated β‐sheet formation for all variants. Self‐assembly rate increased with peptide hydrophobicity; increased molecular volume of the hydrophobic side chain groups did not appear to induce kinetic penalties on self‐assembly rates. Transmission electron microscopy (TEM) imaging indicated variation in fibril morphology as a function of amino acid in the X positions. This study confirms that hydrophobicity of amphipathic Ac‐(XKXE)₂‐NH₂ peptides correlates to self‐assembly propensity and that the hydrophobic core of the resulting nanoribbon bilayers has a significant capacity to accommodate sterically demanding functional groups. These findings provide insight that may be used to guide the exploitation of self‐assembled amphipathic peptides as functional biomaterials.

Solubility Parameters of Amino Acids on Liquid-Liquid Phase Separation and Aggregation of Proteins

The solution properties of amino acids determine the folding, aggregation, and liquid–liquid phase separation (LLPS) behaviors of proteins. Various indices of amino acids, such as solubility, hydropathy, and conformational parameter, describe the behaviors of protein folding and solubility both in vitro and in vivo. However, understanding the propensity of LLPS and aggregation is difficult due to the multiple interactions among different amino acids. Here, the solubilities of aromatic amino acids (SAs) were investigated in solution containing 20 types of amino acids as amino acid solvents. The parameters of SAs in amino acid solvents (PSASs) were varied and dependent on the type of the solvent. Specifically, Tyr and Trp had the highest positive values while Glu and Asp had the lowest. The PSAS values represent soluble and insoluble interactions, which collectively are the driving force underlying the formation of droplets and aggregates. Interestingly, the PSAS of a soluble solvent reflected the affinity between amino acids and aromatic rings, while that of an insoluble solvent reflected the affinity between amino acids and water. These findings suggest that the PSAS can distinguish amino acids that contribute to droplet and aggregate formation, and provide a deeper understanding of LLPS and aggregation of proteins.

Enhancement of protein thermostability by three consecutive mutations using loop-walking method and machine learning

We developed a method to improve protein thermostability, “loop-walking method”. Three consecutive positions in 12 loops of Burkholderia cepacia lipase were subjected to random mutagenesis to make 12 libraries. Screening allowed us to identify L7 as a hot-spot loop having an impact on thermostability, and the P233G/L234E/V235M mutant was found from 214 variants in the L7 library. Although a more excellent mutant might be discovered by screening all the 8000 P233X/L234X/V235X mutants, it was difficult to assay all of them. We therefore employed machine learning. Using thermostability data of the 214 mutants, a computational discrimination model was constructed to predict thermostability potentials. Among 7786 combinations ranked in silico, 20 promising candidates were selected and assayed. The P233D/L234P/V235S mutant retained 66% activity after heat treatment at 60 °C for 30 min, which was higher than those of the wild-type enzyme (5%) and the P233G/L234E/V235M mutant (35%).

IGPRED: Combination of convolutional neural and graph convolutional networks for protein secondary structure prediction

There is a close relationship between the tertiary structure and the function of a protein. One of the important steps to determine the tertiary structure is protein secondary structure prediction (PSSP). For this reason, predicting secondary structure with higher accuracy will give valuable information about the tertiary structure. Recently, deep learning techniques have obtained promising improvements in several machine learning applications including PSSP. In this article, a novel deep learning model, based on convolutional neural network and graph convolutional network is proposed. PSIBLAST PSSM, HHMAKE PSSM, physico-chemical properties of amino acids are combined with structural profiles to generate a rich feature set. Furthermore, the hyper-parameters of the proposed network are optimized using Bayesian optimization. The proposed model IGPRED obtained 89.19%, 86.34%, 87.87%, 85.76%, and 86.54% Q3 accuracies for CullPDB, EVAset, CASP10, CASP11, and CASP12 datasets, respectively.

Identification and Characterization of Species-Specific Severe Acute Respiratory Syndrome Coronavirus 2 Physicochemical Properties

There is an urgent need to elucidate the underlying mechanisms of coronavirus disease (COVID-19) so that vaccines and treatments can be devised. Severe acute respiratory syndrome coronavirus 2 has genetic similarity with bats and pangolin viruses, but a comprehensive understanding of the functions of its proteins at the amino acid sequence level is lacking. A total of 4320 sequences of human and nonhuman coronaviruses was retrieved from the Global Initiative on Sharing All Influenza Data and the National Center for Biotechnology Information. This work proposes an optimization method COVID-Pred with an efficient feature selection algorithm to classify the species-specific coronaviruses based on physicochemical properties (PCPs) of their sequences. COVID-Pred identified a set of 11 PCPs using a support vector machine and achieved 10-fold cross-validation and test accuracies of 99.53% and 97.80%, respectively. These findings could provide key insights into understanding the driving forces during the course of infection and assist in developing effective therapies.

N-semble-based method for identifying Parkinson’s disease genes

Parkinson’s disease (PD) genes identification plays an important role in improving the diagnosis and treatment of the disease. A number of machine learning methods have been proposed to identify disease-related genes, but only few of these methods are adopted for PD. This work puts forth a novel neural network-based ensemble (n-semble) method to identify Parkinson’s disease genes. The artificial neural network is trained in a unique way to ensemble the multiple model predictions. The proposed n-semble method is composed of four parts: (1) protein sequences are used to construct feature vectors using physicochemical properties of amino acid; (2) dimensionality reduction is achieved using the t-Distributed Stochastic Neighbor Embedding (t-SNE) method, (3) the Jaccard method is applied to find likely negative samples from unknown (candidate) genes, and (4) gene prediction is performed with n-semble method. The proposed n-semble method has been compared with Smalter’s, ProDiGe, PUDI and EPU methods using various evaluation metrics. It has been concluded that the proposed n-semble method outperforms the existing gene identification methods over the other methods and achieves significantly higher precision, recall and F Score of 88.9%, 90.9% and 89.8%, respectively. The obtained results confirm the effectiveness and validity of the proposed framework.

Differential effects of serine side chain interactions in amyloid formation by islet amyloid polypeptide

Islet amyloid polypeptide (IAPP), a 37 residue polypeptide, is the main protein component of islet amyloid deposits produced in the pancreas in Type 2 diabetes. Human IAPP contains five serine residues at positions 19, 20, 28, 29, and 34. Models of the IAPP amyloid fibril indicate a structure composed of two closely aligned columns of IAPP monomers with each monomer contributing to two intermolecular β-strands. Ser 19 and Ser 20 are in the partially ordered β-turn region, which links the two strands, whereas Ser 28, Ser 29, and Ser 34 are in the core region of the amyloid fibril. Ser 29 is involved in contacts between the two columns of monomers and is the part of the steric zipper interface. We undertook a study of individual serine substitutions with the hydrophobic isostere 2-aminobutyric acid (2-Abu) to examine the site-specific role of serine side chains in IAPP amyloid formation. All five variants formed amyloid. The Ser 19 to 2-Abu mutant accelerates amyloid formation by a factor of 3 to 4, while the Ser 29 to 2-Abu mutation modestly slows the rate of amyloid formation. 2-Abu replacements at the other sites had even smaller effects. The data demonstrate that the cross-column interactions made by residue 29 are not essential for amyloid formation and also show that cross-strand networks of hydrogen-bonded Ser side chains, so called Ser-ladders, are not required for IAPP amyloid formation. The effect of the Ser 19 to 2-Abu mutant suggests that residues in this region are important for amyloid formation by IAPP.

Amino Acid Encoding Methods for Protein Sequences: A Comprehensive Review and Assessment

As the first step of machine-learning based protein structure and function prediction, the amino acid encoding play a fundamental role in the final success of those methods. Different from the protein sequence encoding, the amino acid encoding can be used in both residue-level and sequence-level prediction of protein properties by combining them with different algorithms. However, it has not attracted enough attention in the past decades, and there are no comprehensive reviews and assessments about encoding methods so far. In this article, we make a systematic classification and propose a comprehensive review and assessment for various amino acid encoding methods. Those methods are grouped into five categories according to their information sources and information extraction methodologies, including binary encoding, physicochemical properties encoding, evolution-based encoding, structure-based encoding, and machine-learning encoding. Then, 16 representative methods from five categories are selected and compared on protein secondary structure prediction and protein fold recognition tasks by using large-scale benchmark datasets. The results show that the evolution-based position-dependent encoding method PSSM achieved the best performance, and the structure-based and machine-learning encoding methods also show some potential for further application, the neural network based distributed representation of amino acids in particular may bring new light to this area. We hope that the review and assessment are useful for future studies in amino acid encoding.

ProteinUnet-An efficient alternative to SPIDER3-single for sequence-based prediction of protein secondary structures

Predicting protein function and structure from sequence remains an unsolved problem in bioinformatics. The best performing methods rely heavily on evolutionary information from multiple sequence alignments, which means their accuracy deteriorates for sequences with a few homologs, and given the increasing sequence database sizes requires long computation times. Here, a single‐sequence‐based prediction method is presented, called ProteinUnet, leveraging an U‐Net convolutional network architecture. It is compared to SPIDER3‐Single model, based on long short‐term memory‐bidirectional recurrent neural networks architecture. Both methods achieve similar results for prediction of secondary structures (both three‐ and eight‐state), half‐sphere exposure, and contact number, but ProteinUnet has two times fewer parameters, 17 times shorter inference time, and can be trained 11 times faster. Moreover, ProteinUnet tends to be better for short sequences and residues with a low number of local contacts. Additionally, the method of loss weighting is presented as an effective way of increasing accuracy for rare secondary structures.

Engineering a Virus-like Particle to Display Peptide Insertions Using an Apparent Fitness Landscape

Peptide insertions in the primary sequence of proteins expand functionality by introducing new binding sequences, chemical handles, or membrane disrupting motifs. With these properties, proteins can be engineered as scaffolds for vaccines or targeted drug delivery vehicles. Virus-like particles (VLPs) are promising platforms for these applications since they are genetically simple, mimic viral structure for cell uptake, and can deliver multiple copies of a therapeutic agent to a given cell. Peptide insertions in the coat protein of VLPs can increase VLP uptake in cells by increasing cell binding, but it is difficult to predict how an insertion affects monomer folding and higher order assembly. To this end, we have engineered the MS2 VLP using a high-throughput technique, called Systematic Mutagenesis and Assembled Particle Selection (SyMAPS). In this work, we applied SyMAPS to investigate a highly mutable loop in the MS2 coat protein to display 9,261 non-native tripeptide insertions. This library generates a discrete map of three amino acid insertions permitted at this location, validates the FG loop as a valuable position for peptide insertion, and illuminates how properties such as charge, flexibility, and hydrogen bonding can interact to preserve or disrupt capsid assembly. Taken together, the results highlight the potential to engineer VLPs in a systematic manner, paving the way to exploring the applications of peptide insertions in biomedically relevant settings.

Alternating Cationic-Hydrophobic Peptide/Peptoid Hybrids: Influence of Hydrophobicity on Antibacterial Activity and Cell Selectivity

The influence of hydrophobicity on antibacterial activity versus the effect on the viability of mammalian cells for peptide/peptoid hybrids was examined for oligomers based on the cationic Lys-like peptoid residue combined with each of 28 hydrophobic amino acids in an alternating sequence. Their relative hydrophobicity was correlated to activity against both Gram-negative and Gram-positive species, human red blood cells, and HepG2 cells. This identified hydrophobic side chains that confer potent antibacterial activity (e. g., MICs of 2-8 μg/mL against E. coli) and low toxicity toward mammalian cells (<10 % hemolysis at 400 μg/mL and IC₅₀ >800 μg/mL for HepG2 viability). Most peptidomimetics retained activity against drug-resistant strains. These findings corroborate the hypothesis that for related peptidomimetics two hydrophobicity thresholds may be identified: i) it should exceed a certain level in order to confer antibacterial activity, and ii) there is an upper limit, beyond which cell selectivity is lost. It is envisioned that once identified for a given subclass of peptide-like antibacterials such thresholds can guide further optimisation.

Peptide-mediated porphyrin based hierarchical complexes for light-to-chemical conversion

We present a new strategy for the biomimetic preparation of integrated photoactive complexes consisting of light harvesting and electron separation/transfer components via the hierarchical assembly of porphyrin and platinum nanoparticles on the surface of short-peptide self-assembled structures. The complexes can catalyze the conversion of visible light energy into chemical energy in the absence of an electron mediator and store it as nicotinamide adenine dinucleotide (NADH). This provides a promising step towards artificial photosystems through precisely controlled interactions of light-capturing agents, electron separators and biomolecules.

HAPPENN is a novel tool for hemolytic activity prediction for therapeutic peptides which employs neural networks

The growing prevalence of resistance to antibiotics motivates the search for new antibacterial agents. Antimicrobial peptides are a diverse class of well-studied membrane-active peptides which function as part of the innate host defence system, and form a promising avenue in antibiotic drug research. Some antimicrobial peptides exhibit toxicity against eukaryotic membranes, typically characterised by hemolytic activity assays, but currently, the understanding of what differentiates hemolytic and non-hemolytic peptides is limited. This study leverages advances in machine learning research to produce a novel artificial neural network classifier for the prediction of hemolytic activity from a peptide's primary sequence. The classifier achieves best-in-class performance, with cross-validated accuracy of [Formula: see text] and Matthews correlation coefficient of 0.71. This innovative classifier is available as a web server at https://research.timmons.eu/happenn , allowing the research community to utilise it for in silico screening of peptide drug candidates for high therapeutic efficacies.

Deep Dive into Machine Learning Models for Protein Engineering

Protein redesign and engineering has become an important task in pharmaceutical research and development. Recent advances in technology have enabled efficient protein redesign by mimicking natural evolutionary mutation, selection, and amplification steps in the laboratory environment. For any given protein, the number of possible mutations is astronomical. It is impractical to synthesize all sequences or even to investigate all functionally interesting variants. Recently, there has been an increased interest in using machine learning to assist protein redesign, since prediction models can be used to virtually screen a large number of novel sequences. However, many state-of-the-art machine learning models, especially deep learning models, have not been extensively explored. Moreover, only a small selection of protein sequence descriptors has been considered. In this work, the performance of prediction models built using an array of machine learning methods and protein descriptor types, including two novel, single amino acid descriptors and one structure-based three-dimensional descriptor, is benchmarked. The predictions were evaluated on a diverse collection of public and proprietary data sets, using a variety of evaluation metrics. The results of this comparison suggest that Convolution Neural Network models built with amino acid property descriptors are the most widely applicable to the types of protein redesign problems faced in the pharmaceutical industry.

Natural amino acids as potential chelators for soil remediation

The soils contaminated by toxic metals are often remediated using EDTA and similar non-biodegradable chelators. Most chelators are in fact synthetic amino acid derivatives, whereas natural proteinogenic amino acids (PAAs) have not been systematically explored as remediation agents, despite their well-known metal chelating abilities and environmental benefits. Our study represents a comprehensive research exploring 16 structurally and functionally different PAAs as potential remediating agents, applied to 3 different heavy metal-contaminated samples. The study was mostly focused on extracting Cd, Cu, Ni, and Zn. The reaction parameters were screened and optimized. It was found that the efficiencies of extracting Cu, Ni, and Zn by Threonine, Aspartic acid and Histidine were comparable to those by EDTA, whereas non-polar side chain-containing PAAs demonstrated consistently lower PTM extraction rates compared to other PAAs. The sulfur-containing Cysteine appeared to be efficient to extract Cd (to some extent), Ni and Zn, but not Cu, due to chemical reasons. The structure-functional correlations were identified, described, and found to be independent on the specific samples. Possible molecular mechanisms of metal extraction from soils by PAAs are discussed. In contrast to EDTA, the soil-essential elements are almost not extracted by PAAs. This important feature of the PAAs, along with their availability, observed selectivity, competitive efficiency, non-toxicity and even fertilizing properties, make them particularly soil-friendly, and thus, potentially applicable chelators in certain remediation processes.

Protein Secondary Structure Prediction: A Review of Progress and Directions

Background:

Over the last few decades, a search for the theory of protein folding has grown into a full-fledged research field at the intersection of biology, chemistry and informatics. Despite enormous effort, there are still open questions and challenges, like understanding the rules by which amino acid sequence determines protein secondary structure.

Objective:

In this review, we depict the progress of the prediction methods over the years and identify sources of improvement.

Methods:

The protein secondary structure prediction problem is described followed by the discussion on theoretical limitations, description of the commonly used data sets, features and a review of three generations of methods with the focus on the most recent advances. Additionally, methods with available online servers are assessed on the independent data set.

Results:

The state-of-the-art methods are currently reaching almost 88% for 3-class prediction and 76.5% for an 8-class prediction.

Conclusion:

This review summarizes recent advances and outlines further research directions.

The Role of Hydrophobicity in the Stability and pH-Switchability of (RXDX)4 and Coumarin-(RXDX)4 Conjugate β-Sheets

pH-Switchable, self-assembling materials are of interest in biological imaging and sensing applications. Here we propose that combining the pH-switchability of RXDX (X = Ala, Val, Leu, Ile, Phe) peptides and the optical properties of coumarin creates an ideal candidate for these materials. This suggestion is tested with a thorough set of all-atom molecular dynamics simulations. We first investigate the dependence of pH-switchabiliy on the identity of the hydrophobic residue, X, in the bare (RXDX)₄ systems. Increasing the hydrophobicity stabilizes the fiber which, in turn, reduces the pH-switchabilty of the system. This behavior is found to be somewhat transferable to systems in which a single hydrophobic residue is replaced with a coumarin containing amino acid. In this case, conjugates with X = Ala are found to be unstable at both pHs, while conjugates with X = Val, Leu, Ile, and Phe are found to form stable β-sheets at least at neutral pH. The coumarin-(RFDF)₄ conjugate is found to have the largest relative entropy value of 0.884 ± 0.001 between neutral and acidic coumarin ordering distributions. Thus, we posit that coumarin-(RFDF)₄ containing peptide sequences are ideal candidates for pH-sensing bioelectronic materials.

VarSite: Disease variants and protein structure

VarSite is a web server mapping known disease-associated variants from UniProt and ClinVar, together with natural variants from gnomAD, onto protein 3D structures in the Protein Data Bank. The analyses are primarily image-based and provide both an overview for each human protein, as well as a report for any specific variant of interest. The information can be useful in assessing whether a given variant might be pathogenic or benign. The structural annotations for each position in the protein include protein secondary structure, interactions with ligand, metal, DNA/RNA, or other protein, and various measures of a given variant's possible impact on the protein's function. The 3D locations of the disease-associated variants can be viewed interactively via the 3dmol.js JavaScript viewer, as well as in RasMol and PyMOL. Users can search for specific variants, or sets of variants, by providing the DNA coordinates of the base change(s) of interest. Additionally, various agglomerative analyses are given, such as the mapping of disease and natural variants onto specific Pfam or CATH domains. The server is freely accessible to all at: https://www.ebi.ac.uk/thornton-srv/databases/VarSite.

Relevance of Machine Learning Techniques and Various Protein Features in Protein Fold Classification: A Review

Background:

Protein fold prediction is a fundamental step in Structural Bioinformatics. The tertiary structure of a protein determines its function and to predict its tertiary structure, fold prediction serves an important role. Protein fold is simply the arrangement of the secondary structure elements relative to each other in space. A number of studies have been carried out till date by different research groups working worldwide in this field by using the combination of different benchmark datasets, different types of descriptors, features and classification techniques.

Objective:

In this study, we have tried to put all these contributions together, analyze their study and to compare different techniques used by them.

Methods:

Different features are derived from protein sequence, its secondary structure, different physicochemical properties of amino acids, domain composition, Position Specific Scoring Matrix, profile and threading techniques.

Conclusion:

Combination of these different features can improve classification accuracy to a large extent. With the help of this survey, one can know the most suitable feature/attribute set and classification technique for this multi-class protein fold classification problem.