Field Of Computational Biology Research Articles

MFTrans: A multi-feature transformer network for protein secondary structure prediction

In the rapidly evolving field of computational biology, accurate prediction of protein secondary structures is crucial for understanding protein functions, facilitating drug discovery, and advancing disease diagnostics. In this paper, we propose MFTrans, a deep learning-based multi-feature fusion network aimed at enhancing the precision and efficiency of Protein Secondary Structure Prediction (PSSP). This model employs a Multiple Sequence Alignment (MSA) Transformer in combination with a multi-view deep learning architecture to effectively capture both global and local features of protein sequences. MFTrans integrates diverse features generated by protein sequences, including MSA, sequence information, evolutionary information, and hidden state information, using a multi-feature fusion strategy. The MSA Transformer is utilized to interleave row and column attention across the input MSA, while a Transformer encoder and decoder are introduced to enhance the extracted high-level features. A hybrid network architecture, combining a convolutional neural network with a bidirectional Gated Recurrent Unit (BiGRU) network, is used to further extract high-level features after feature fusion. In independent tests, our experimental results show that MFTrans has superior generalization ability, outperforming other state-of-the-art PSSP models by 3 % on average on public benchmarks including CASP12, CASP13, CASP14, TEST2016, TEST2018, and CB513. Case studies further highlight its advanced performance in predicting mutation sites. MFTrans contributes significantly to the protein science field, opening new avenues for drug discovery, disease diagnosis, and protein.

BioProtIS: Streamlining protein-ligand interaction pipeline for analysis in genomic and transcriptomic exploration

The identification of protein-ligand interactions plays a pivotal role in elucidating biological processes and discovering potential bioproducts. Harnessing the capabilities of computational methods in drug discovery, we introduce an innovative Inverted Virtual Screening (IVS) pipeline. This pipeline Integrated molecular dynamics and docking analyses to ensure that protein structures are not only energetically favorable but also representative of stable conformations. The primary objective of this pipeline is to automate and streamline the analysis of protein-ligand interactions at both genomic and transcriptomic scales. In the contemporary post-genomic era, high-throughput computational screening for bioproducts, biological systems, and therapeutic drugs has become a cornerstone practice. This approach offers the promise of cost-effectiveness, time efficiency, and optimization of laboratory work. Nevertheless, a notable deficiency persists in the availability of efficient pipelines capable of automating the virtual screening process, seamlessly integrating input and output, and leveraging the full potential of open-source tools. To bridge this critical gap, we have developed a versatile pipeline known as BioProtIS. This tool seamlessly integrates a suite of state-of-the-art tools, including Modeller, AlphaFold, Gromacs, FPOCKET, and AutoDock Vina, thus facilitating the streamlined docking of ligands with an expansive repertoire of proteins sourced from genomes and transcriptomes, and substrates. To assess the pipeline's performance, we employed the transcriptomes of Cereus jamacaru (a cactus species) and Aspisoma lineatum (firefly), along with the genome of Homo sapiens. This integration not only improves the accuracy of ligand-protein interactions by minimizing replicability deviations but also optimizes the discovery process by enabling the simultaneous evaluation of multiple substrates. Furthermore, our pipeline accommodates distinct testing scenarios, such as blind docking or site-specific targeting, which are invaluable in applications ranging from drug repositioning to the exploration of new allosteric binding sites and toxicity assessments. BioProtIS has been designed with modularity at its core. This inherent flexibility empowers users to make custom modifications directly within the source code, tailoring the pipeline to their specific research needs. Moreover, it lays the foundation for seamless integration of diverse docking algorithms in future iterations, promising ongoing advancements in the field of computational biology. This pipeline is available for free distribution and can be download at: https://github.com/BBMDO/BioProtIS.

Inferring a directed acyclic graph of phenotypes from GWAS summary statistics.

Estimating phenotype networks is a growing field in computational biology. It deepens the understanding of disease etiology and is useful in many applications. In this study, we present a method that constructs a phenotype network by assuming a Gaussian linear structure model embedding a directed acyclic graph (DAG). We utilize genetic variants as instrumental variables and show how our method only requires access to summary statistics from a genome-wide association study (GWAS) and a reference panel of genotype data. Besides estimation, a distinct feature of the method is its summary statistics-based likelihood ratio test on directed edges. We applied our method to estimate a causal network of 29 cardiovascular-related proteins and linked the estimated network to Alzheimer's disease (AD). A simulation study was conducted to demonstrate the effectiveness of this method. An R package sumdag implementing the proposed method, all relevant code, and a Shiny application are available.

Positional difference and Frequency (PdF) based alignment-free technique for genome sequence comparison

In the field of computational biology, genome sequence comparison among different species is essential and has applications in both the research and scientific fields. Owing to the lengthy processing time and large number of data sets, the alignment-based approaches are unsuitable and ineffective. Therefore, alignment-free techniques have obtained popularity for acquiring proper sequence clustering and evolutionary relationship among species. In this paper, a complete bipartite graph based Positional difference and Frequency (PdF) vector descriptor is introduced. Positional difference and Frequency, two parameters, are applied to the genome sequence to create a 16- dimensional vector descriptor using the di-nucleotide representation of genome sequence. Subsequently, a distance matrix is calculated to construct the phylogenetic trees for different data sets of mammals and viruses. The achieved outcomes are compared with the phylogenetic trees of the earlier methods viz. the FFP method, the ClustalW method, the MEV method, the PCNV method and the FIS method. In most instances, the proposed method produces more precise outcomes than the preceding techniques and has potential for genome sequence comparison on both the equal and unequal length of data-sets. Communicated by Ramaswamy H. Sarma

Big Data Analysis in Computational Biology and Bioinformatics.

Advancements in high-throughput technologies, genomics, transcriptomics, and metabolomics play an important role in obtaining biological information about living organisms. The field of computational biology and bioinformatics has experienced significant growth with the advent of high-throughput sequencing technologies and other high-throughput techniques. The resulting large amounts of data present both opportunities and challenges for data analysis. Big data analysis has become essential for extracting meaningful insights from the massive amount of data. In this chapter, we provide an overview of the current status of big data analysis in computational biology and bioinformatics. We discuss the various aspects of big data analysis, including data acquisition, storage, processing, and analysis. We also highlight some of the challenges and opportunities of big data analysis in this area of research. Despite the challenges, big data analysis presents significant opportunities like development of efficient and fast computing algorithms for advancing our understanding of biological processes, identifying novel biomarkers for breeding research and developments, predicting disease, and identifying potential drug targets for drug development programs.

A Review of Computational Approach for S-system-based Modeling of Gene Regulatory Network.

Inference of gene regulatory network (GRN) from time series microarray data remains as a fascinating task for computer science researchers to understand the complex biological process that occurred inside a cell. Among the different popular models to infer GRN, S-system is considered as one of the promising non-linear mathematical tools to model the dynamics of gene expressions, as well as to infer the GRN. S-system is based on biochemical system theory and power law formalism. By observing the value of kinetic parameters of S-system model, it is possible to extract the regulatory relationships among genes. In this review, several existing intelligent methods that were already proposed for inference of S-system-based GRN are explained. It is observed that finding out the most suitable and efficient optimization technique for the accurate inference of all kinds of networks, i.e., in-silico, in-vivo, etc., with less computational complexity is still an open research problem to all. This paper may help the beginners or researchers who want to continue their research in the field of computational biology and bioinformatics.

Rational peptide design for targeting cancer cell invasion.

Cell invasion is an important process in cancer progression and recurrence. Invasion and implantation of cancer cells from their original place to other tissues, by disabling vital organs, challenges the treatment of cancer patients. Given the importance of the matter, many molecular treatments have been developed to inhibit cancer cell invasion. Because of their low production cost and ease of production, peptides are valuable therapeutic molecules for inhibiting cancer cell invasion. In recent years, advances in the field of computational biology have facilitated the design of anti-cancer peptides. In our investigation, using computational biology approaches such as evolutionary analysis, residue scanning, protein-peptide interaction analysis, molecular dynamics, and free energy analysis, our team designed a peptide library with about 100 000 candidates based on A6 (acetyl-KPSSPPEE-amino) sequence which is an anti-invasion peptide. During computational studies, two of the designed peptides that give the highest scores and showed the greatest sequence similarity to A6 were entered into the experimental analysis workflow for further analysis. In experimental analysis steps, the anti-metastatic potency and other therapeutic effects of designed peptides were evaluated using MTT assay, RT-qPCR, zymography analysis, and invasion assay. Our study disclosed that the IK1 (acetyl-RPSFPPEE-amino) peptide, like A6, has great potency to inhibit the invasion of cancer cells.

Computer Aided Reverse Vaccinology: A Game-changer Approach for Vaccine Development.

One of the most dynamic approaches in biotechnology is reverse vaccinology, which plays a huge role in today's developing vaccines. It has the capability of exploring and identifying the most potent vaccine candidate in a limited period of time. The first successful novel approach of reverse vaccinology was observed in Neisseria meningitidis serogroup B, which has revolutionised the whole field of computational biology. In this review, we have summarized the application of reverse vaccinology for different infectious diseases, discussed epitope prediction and various available bioinformatic tools, and explored the advantages, limitations and necessary elements of this approach. Some of the modifications in the reverse vaccinology approach, like pan-genome and comparative reverse vaccinology, are also outlined. Vaccines for illnesses like AIDS and hepatitis C have not yet been developed. Computer Aided Reverse vaccinology has the potential to be a game-changer in this area. The use of computational tools, pipelines and advanced soft-computing methods, such as artificial intelligence and deep learning, and exploitation of available omics data in integration have paved the way for speedy and effective vaccine designing. Is reverse vaccinology a viable option for developing vaccines against such infections, or is it a myth? Vaccine development gained momentum after the spread of various infections, resulting in numerous deaths; these vaccines are developed using the traditional technique, which includes inactivated microorganisms. As a result, reverse vaccinology may be a far superior technique for creating an effective vaccine.

PGAGP: Predicting pathogenic genes based on adaptive network embedding algorithm.

The study of disease-gene associations is an important topic in the field of computational biology. The accumulation of massive amounts of biomedical data provides new possibilities for exploring potential relations between diseases and genes through computational strategy, but how to extract valuable information from the data to predict pathogenic genes accurately and rapidly is currently a challenging and meaningful task. Therefore, we present a novel computational method called PGAGP for inferring potential pathogenic genes based on an adaptive network embedding algorithm. The PGAGP algorithm is to first extract initial features of nodes from a heterogeneous network of diseases and genes efficiently and effectively by Gaussian random projection and then optimize the features of nodes by an adaptive refining process. These low-dimensional features are used to improve the disease-gene heterogenous network, and we apply network propagation to the improved heterogenous network to predict pathogenic genes more effectively. By a series of experiments, we study the effect of PGAGP's parameters and integrated strategies on predictive performance and confirm that PGAGP is better than the state-of-the-art algorithms. Case studies show that many of the predicted candidate genes for specific diseases have been implied to be related to these diseases by literature verification and enrichment analysis, which further verifies the effectiveness of PGAGP. Overall, this work provides a useful solution for mining disease-gene heterogeneous network to predict pathogenic genes more effectively.

Hardness and approximation of multiple sequence alignment with column score

Multiple Sequence Alignment (MSA for short) is a well known problem in the field of computational biology. In order to evaluate the quality of a solution, many different scoring functions have been introduced, the most widely used being the Sum-of-Pairs score (SP-score). It is known that computing the best MSA under the SP-score measure is NP-hard.In this paper, we introduce a variant of the Column score (defined in Thompson et al. 1999), which we refer to as Selective Column score: Given a symbol a∈Σ, the score of the i-th column is one if and only if all symbols of the same column are a, and otherwise zero. The a-column score of an alignment is then the number of columns made of only character a.We show that finding the optimal MSA under the Selective Column Score is NP-hard for all alphabets of size |Σ|≥2, and that the associated maximization problem is poly-APX-hard.We also give an approximation algorithm that almost matches the inapproximability bound.

New Approaches for String Site-Removal Problem

The string site-removal problem (SSRP) originates from the field of computational biology and is known to be NP-hard. The problem is defined on the set of strings of equal length and an integer k, and its goal is to find the largest subset of the strings that become identical after removing at most k sites (columns). This paper proposes two new approaches to SSRP. The first is an exact algorithm that uses the quadratically constrained program (QCP). The second approach combines QCP and genetic algorithm (GA). In a series of experiments, we showed that our exact program performs faster on average than the state-of-the-art method (Integer Linear Programming, ILP). Regarding heuristic approaches, our new proposal gives the identical results as the state-of-the-art reference method (Large Neighborhood Search, LNS) with the same CPU time budget. All source codes and benchmark instances we made publicly available via GitHub.

Antiproliferative Activity Predictor: A New Reliable In Silico Tool for Drug Response Prediction against NCI60 Panel.

In vitro antiproliferative assays still represent one of the most important tools in the anticancer drug discovery field, especially to gain insights into the mechanisms of action of anticancer small molecules. The NCI-DTP (National Cancer Institute Developmental Therapeutics Program) undoubtedly represents the most famous project aimed at rapidly testing thousands of compounds against multiple tumor cell lines (NCI60). The large amount of biological data stored in the National Cancer Institute (NCI) database and many other databases has led researchers in the fields of computational biology and medicinal chemistry to develop tools to predict the anticancer properties of new agents in advance. In this work, based on the available antiproliferative data collected by the NCI and the manipulation of molecular descriptors, we propose the new in silico Antiproliferative Activity Predictor (AAP) tool to calculate the GI50 values of input structures against the NCI60 panel. This ligand-based protocol, validated by both internal and external sets of structures, has proven to be highly reliable and robust. The obtained GI50 values of a test set of 99 structures present an error of less than ±1 unit. The AAP is more powerful for GI50 calculation in the range of 4-6, showing that the results strictly correlate with the experimental data. The encouraging results were further supported by the examination of an in-house database of curcumin analogues that have already been studied as antiproliferative agents. The AAP tool identified several potentially active compounds, and a subsequent evaluation of a set of molecules selected by the NCI for the one-dose/five-dose antiproliferative assays confirmed the great potential of our protocol for the development of new anticancer small molecules. The integration of the AAP tool in the free web service DRUDIT provides an interesting device for the discovery and/or optimization of anticancer drugs to the medicinal chemistry community. The training set will be updated with new NCI-tested compounds to cover more chemical spaces, activities, and cell lines. Currently, the same protocol is being developed for predicting the TGI (total growth inhibition) and LC50 (median lethal concentration) parameters to estimate toxicity profiles of small molecules.

Environmental microbiology going computational-Predictive ecology and unpredicted discoveries.

The fields of microbial ecology and environmental microbiology are producing loads of data, mainly nucleic acid sequence data due to the extensive use of amplicon sequencing and metagenomics, and an increasing use of transcriptomics. To increase our understanding of microorganisms in terrestrial ecosystems, multiple, concerted efforts to collect large numbers of samples for analyses of microbial communities were initiated already more than 15 years ago (Fierer & Jackson, 2006; Lozupone & Knight, 2007) but have really exploded the last years, with The Earth Microbiome Project Consortium being one of the first major endeavours for bacteria across all biomes (Thompson et al., 2017) and the work by Tedersoo et al. (2014) for soil fungi. The majority of the investigations have a biogeography focus based on a single sampling occasion and the word ‘global’ is frequently used in the titles of these soil microbial catalogues and surveys (Bahram et al., 2018; Delgado-Baquerizo et al., 2018; Gobbi et al., 2022). Similar efforts have been done for many other biomes. Although largely descriptive, they have contributed to a better understanding of microbial diversity and the distribution of microbial taxa and their functions at an unprecedented spatial scale. Further, correlative analyses have indicted direct or indirect drivers of the observed patterns as well as the role of microbial communities for ecosystem functioning (Bahram et al., 2018; Delgado-Baquerizo et al., 2020; Garland et al., 2021). The massive amount of complex data is not only an opportunity but also a major challenge when it comes to meaningful interpretation. The field of computational biology, being the intersection of computer science and biology, is rapidly expanding and developing new methods for this purpose. Artificial intelligence (AI), including machine learning (ML) and to some extent also deep learning (DL) methods are promising for dealing with big data in microbial ecology and environmental microbiology (Ghannam & Techtmann, 2021; McElhinney et al., 2022). Especially ML approaches are increasingly adopted by ecologists and many of these methods will soon become routine tools for analyses of complex microbial omics data. They can be used to categorize and finds patterns in uncategorized data as well as analyse data that we know how to categorize. There are several advantages to using ML methods in microbiome studies, for example, they can deal with non-linear relationships, make better use of the full depth of high-dimensional data, and can be used to build predictive models based on environmental and community data. Predictive modelling is very attractive in microbial ecology. Among the ML methods, random forests have become frequently applied in microbiome studies in the last decade (Jones et al., 2014; Ryo & Rillig, 2017). It is predominantly used for the identification of the best predictors for a given response variable and has for example been used to rank the environmental variables determining the major microbial phyla in wetlands (Bahram et al., 2022) and the diversity of ammonia oxidizing archaea across European soils (Saghaï et al., 2022), as well as the relative importance of biotic and abiotic controls of nitrous oxide emissions from agricultural soils (Jones et al., 2022). Random forest modelling can be very useful when studying remote areas that are difficult to sample, as exemplified by climate projections on microbial communities in the Antarctic Ocean (Tonelli et al., 2021). RF models can also show how predictions change over the range of each individual predictor variable, thereby giving the possibility to identify thresholds or tipping points (Apley & Zhu, 2020; Saghaï et al., 2022). Already in 2012, artificial neural networks were used to incorporate interactions among community members in models for predictions of microbial community composition in time and space based on environmental data (Larsen et al., 2012). A similar approach was used to predict the maize rhizosphere community at different plant development stages or growth conditions (García-Jiménez et al., 2021). This type of approach can potentially assist in the microbiome engineering of important crops. However, with sequencing costs being relatively cheap, there is an increasing interest in using AI and microbiome data for microbiome-based diagnostics as a means to address environmental challenges and advance management practices (McElhinney et al., 2022). Two recent examples of the latter are the use of soil microbiome data to predict the propensity for specific plant diseases in agriculture (Yuan et al., 2020) and soil health metrics (Wilhelm et al., 2022), which can be laborious and expensive to measure. Combining ML and microbiome data has further shown promising in environmental monitoring, tracing of contaminants and predictions of environmental quality (Sperlea et al., 2022; Techtmann & Hazen, 2016; Wheeler, 2019), which allows us to move away from indicator taxa or microbial biomarkers and instead use the full breath of information encompassed by the microbial community in a given site or sample. The large amounts of genetic data and corresponding meta-data generated in microbiome studies are real treasures, especially when it comes to metagenomes and metatranscriptomes, and only a fraction of the information available has been explored. This data can be used for meta-analyses to increase the scale of the study, but more importantly, it can be used to address other questions than those posed by the researchers that collected the original data. Making use of already published genome or sequence data in microbial ecology is not a new idea (Jones & Hallin, 2010) but now we have increasing possibilities to mine extremely large data sets (Coelho et al., 2022). Even more exciting are the possibilities to combine different types of data and information to go beyond the microbiome data. Integration of knowledge from diverse fields of research and the combination of microbiome data with other data from different sources have the potential to result in unexpected and unpredictable results, as well as new discoveries. A recent example of re-using and combining data is the work by Ke et al. (2022), who reanalyzed data in published datasets on the effects of pesticide application on soil microbial communities combined with information on the physical and chemical properties of the pesticides. By developing a ML model, they were able to show that physical pesticide properties largely explain the ecological impact of the pesticide. This information can guide the design of pesticide molecules to minimize environmental risk. In the field of precision agriculture, researchers have proposed the integration of AI and nanotechnology with disparate datasets to enable the design of nanoscale agrochemicals for sustainable food production (Zhang et al., 2021). In another study, geographic and meteorological data as well plant-traits, land-use type and microbial community data were used in a ML-based prediction of grassland degradation, which is a multi-factorial phenomenon not easily captured by a few variables (Yan et al., 2022). Combining datasets and using computational approaches can also be used to develop new diagnostic tools. For example, de Andrade et al. (2021) suggest the development of a soil quality index based on soil microbiome data, crop productivity and a range of abiotic environmental factors to improve crop production systems using AI. Data-driven research relying on large, multiple, complex datasets and computational methods and capacity, as exemplified above, indicates a new paradigm in microbial ecology, and ecology in general (McCallen et al., 2019). We can anticipate new insights, similar to the leaps taken after advanced bioinformatics and multi-omics approaches became an integral part of microbial ecology research. Microbial ecology and environmental microbiology will follow the trajectory in life sciences and become increasingly computationally demanding, focusing on larger and also more complex sets of information. We are already seeing the laboratories being sparsely populated while students, postdocs, and researchers spend increasing amount of time in front of their computers organizing and analysing data. My crystal ball says that a shift towards a data-driven rather than an experimental-driven and data generating science, that depends on complex, big data, and advanced technologies, will be a game changer in microbial ecology and environmental microbiology. This development is already putting pressure on management, storage and sharing of data. Data-driven microbial ecology research where different types of data are combined to consider the multidimensionality of ecosystems further suggests that students and researchers not only need to enhance their computational skills, but also skills in working interdisciplinary. Nevertheless, important discoveries should ideally be followed by experimental approaches to test hypothesis, determine causal relationships, and verify mechanisms. Already, experimental validation is definitely a bottleneck to close the circle in microbial ecology research and, although my crystal ball is a bit hazy here, it looks like this will become an even greater bottleneck in the era of big data and data-drivenresearch in microbial ecology. This work was supported by the Swedish University of Agricultural Sciences (senior career grant 2019-2024) and the Swedish Research Council (grant 2016-03551). There is no data associated with this article.

CMash: fast, multi-resolution estimation of k-mer-based Jaccard and containment indices.

Motivation K-mer-based methods are used ubiquitously in the field of computational biology. However, determining the optimal value of k for a specific application often remains heuristic. Simply reconstructing a new k-mer set with another k-mer size is computationally expensive, especially in metagenomic analysis where datasets are large. Here, we introduce a hashing-based technique that leverages a kind of bottom-m sketch as well as a k-mer ternary search tree (KTST) to obtain k-mer-based similarity estimates for a range of k values. By truncating k-mers stored in a pre-built KTST with a large value, we can simultaneously obtain k-mer-based estimates for all k values up to kmax. This truncation approach circumvents the reconstruction of new k-mer sets when changing k values, making analysis more time and space-efficient.ResultsWe derived the theoretical expression of the bias factor due to truncation. And we showed that the biases are negligible in practice: when using a KTST to estimate the containment index between a RefSeq-based microbial reference database and simulated metagenome data for 10 values of k, the running time was close to 10× faster compared to a classic MinHash approach while using less than one-fifth the space to store the data structure.Availability and implementationA python implementation of this method, CMash, is available at https://github.com/dkoslicki/CMash. The reproduction of all experiments presented herein can be accessed via https://github.com/KoslickiLab/CMASH-reproducibles.Supplementary information Supplementary data are available at Bioinformatics online.

Michael Waterman's Contributions to Computational Biology and Bioinformatics.

On the occasion of Dr. Michael Waterman's 80th birthday, we review his major contributions to the field of computational biology and bioinformatics including the famous Smith-Waterman algorithm for sequence alignment, the probability and statistics theory related to sequence alignment, algorithms for sequence assembly, the Lander-Waterman model for genome physical mapping, combinatorics and predictions of ribonucleic acid structures, word counting statistics in molecular sequences, alignment-free sequence comparison, and algorithms for haplotype block partition and tagSNP selection related to the International HapMap Project. His books Introduction to Computational Biology: Maps, Sequences and Genomes for graduate students and Computational Genome Analysis: An Introduction geared toward undergraduate students played key roles in computational biology and bioinformatics education. We also highlight his efforts of building the computational biology and bioinformatics community as the founding editor of the Journal of Computational Biology and a founding member of the International Conference on Research in Computational Molecular Biology (RECOMB).

Prior knowledge facilitates low homologous protein secondary structure prediction with DSM distillation.

Protein secondary structure prediction (PSSP) is one of the fundamental and challenging problems in the field of computational biology. Accurate PSSP relies on sufficient homologous protein sequences to build the multiple sequence alignment (MSA). Unfortunately, many proteins lack homologous sequences, which results in the low quality of MSA and poor performance. In this article, we propose the novel dynamic scoring matrix (DSM)-Distil to tackle this issue, which takes advantage of the pretrained BERT and exploits the knowledge distillation on the newly designed DSM features. Specifically, we propose the DSM to replace the widely used profile and PSSM (position-specific scoring matrix) features. DSM could automatically dig for the suitable feature for each residue, based on the original profile. Namely, DSM-Distil not only could adapt to the low homologous proteins but also is compatible with high homologous ones. Thanks to the dynamic property, DSM could adapt to the input data much better and achieve higher performance. Moreover, to compensate for low-quality MSA, we propose to generate the pseudo-DSM from a pretrained BERT model and aggregate it with the original DSM by adaptive residue-wise fusion, which helps to build richer and more complete input features. In addition, we propose to supervise the learning of low-quality DSM features using high-quality ones. To achieve this, a novel teacher-student model is designed to distill the knowledge from proteins with high homologous sequences to that of low ones. Combining all the proposed methods, our model achieves the new state-of-the-art performance for low homologous proteins. Compared with the previous state-of-the-art method 'Bagging', DSM-Distil achieves an improvement about 5% and 7.3% improvement for proteins with MSA count ≤30 and extremely low homologous cases, respectively. We also compare DSM-Distil with Alphafold2 which is a state-of-the-art framework for protein structure prediction. DSM-Distil outperforms Alphafold2 by 4.1% on extremely low-quality MSA on 8-state secondary structure prediction. Moreover, we release a large-scale up-to-date test dataset BC40 for low-quality MSA structure prediction evaluation. BC40 dataset: https://drive.google.com/drive/folders/15vwRoOjAkhhwfjDk6-YoKGf4JzZXIMC. HardCase dataset: https://drive.google.com/drive/folders/1BvduOr2b7cObUHy6GuEWk-aUkKJgzTUv. Code: https://github.com/qinwang-ai/DSM-Distil.

DDREL: From drug-drug relationships to drug repurposing

Analyzing the relationships among various drugs is an essential issue in the field of computational biology. Different kinds of informative knowledge, such as drug repurposing, can be extracted from drug-drug relationships. Scientific literature represents a rich source for the retrieval of knowledge about the relationships between biological concepts, mainly drug-drug, disease-disease, and drug-disease relationships. In this paper, we propose DDREL as a general-purpose method that applies deep learning on scientific literature to automatically extract the graph of syntactic and semantic relationships among drugs. DDREL remarkably outperforms the existing human drug network method and a random network respected to average similarities of drugs’ anatomical therapeutic chemical (ATC) codes. DDREL is able to shed light on the existing deficiency of the ATC codes in various drug groups. From the DDREL graph, the history of drug discovery became visible. In addition, drugs that had repurposing score 1 (diflunisal, pargyline, fenofibrate, guanfacine, chlorzoxazone, doxazosin, oxymetholone, azathioprine, drotaverine, demecarium, omifensine, yohimbine) were already used in additional indication. The proposed DDREL method justifies the predictive power of textual data in PubMed abstracts. DDREL shows that such data can be used to 1- Predict repurposing drugs with high accuracy, and 2- Reveal existing deficiencies of the ATC codes in various drug groups.

Multivariate Data Exploration Through Coordinated Views

Many fields of study still face the challenges inherent to the analysis of complex multidimensional datasets, such as the field of computational biology, whose research of infectious diseases must contend with large protein-protein interaction networks with thousands of genes that vary in expression values over time. In this paper, we explore the visualization of multivariate data through CroP, a data visualization tool with a coordinated multiple views framework where users can adapt the workspace to different problems through flexible panels. In particular, we focus on the visualization of relational and temporal data, the latter being represented through layouts that distort timelines to represent the fluctuations of values across complex datasets, creating visualizations that highlight significant events and patterns. Moreover, CroP provides various layouts and functionalities to not only highlight relationships between different variables, but also dig-down into discovered patterns in order to better understand their sources and their effects. These methods are demonstrated through multiple experiments with diverse multivariate datasets, with a focus on gene expression time-series datasets. In addition to a discussion of our results, we also validate CroP through model and interface tests performed with participants from both the fields of information visualization and computational biology.

Combination of classical and statistical approaches to enhance the fermentation conditions and increase the yield of Lipopeptide(s) by Pseudomonas sp. OXDC12: its partial purification and determining antifungal property.

Around 200 different lipopeptides (LPs) have been identified to date, most of which are produced via Bacillus and Pseudomonas species. The clinical nature of the lipopeptide (LP) has led to a big surge in its research. They show antimicrobial and antitumor activities due to which mass-scale production and purification of LPs are beneficial. Response surface methodology (RSM) approach has emerged as an alternative in the field of computational biology for optimizing the reaction parameters using statistical models. In the present study, Pseudomonas sp. strain OXDC12 was used for production and partial purification of LPs using Thin Layer Chromatography (TLC). The main goal of the study was to increase the overall yield of LPs by optimizing the different variables in the fermentation broth. This was achieved using a combination of one factor at a time (OFAT) and response surface methodology (RSM) approaches. OFAT technique was used to optimize the necessary parameters and was followed by the creation of statistical models (RSM) to optimize the remaining variables. Maximum mycelial growth inhibition (%) against the fungus Mucor sp. was 61.3% for LP. Overall, the combination of both OFAT and RSM helped in increasing the LPs yield by 3 folds from 367mg/L to 1169mg/L.

Big Data and its Future in Computational Biology: A Literature Review

Big data technologies are used majorly for computational biology research nowadays due to the emerge of massive amounts of biological data that have been generated and collected at an unprecedented speed and scale and with different structures. As an example, the processing of billions of DNA sequence data per day represents the new generation of sequencing technologies. It is expected that the cost of acquiring and analyzing biomedical data will decrease with the assistance of technology upgrades in the computational biology field. The application of big data in health care is a fast-growing field, with many new discoveries and methodologies. In this study, an overview of recent developments and technologies in big data in the computational biology field. Moreover, in this study, we summarize and highlight the challenges, gaps and opportunities to improve and advance big data applications in computational biology. Also, this review presents a useful starting point for the application of Big Data in future Bioinformatics research. Finally, some of the Big Data sources in computational biology that help in applying these new emerging technologies are mentioned.