Current Bioinformatics - Volume 17, Issue 7, 2022

Background: Single-cell technologies provide unprecedented opportunities to study heterogeneity of molecular mechanisms. In particular, single-cell RNA-sequence data have been successfully used to infer gene regulatory networks with stochastic expressions. However, there are still substantial challenges in measuring the relationships between genes and selecting the important genetic regulations. Objective: This prospective provides a brief review of effective methods for the inference of gene regulatory networks. Methods: We concentrate on two types of inference methods, namely the model-free methods and mechanistic methods for constructing gene networks. Results: For the model-free methods, we mainly discuss two issues, namely the measures for quantifying gene relationship and criteria for selecting significant connections between genes. The issue for mechanistic methods is different mathematical models to describe genetic regulations accurately. Conclusions: We advocate the development of ensemble methods that combine two or more methods together.

Background: In biology, the translation of genetic information to its corresponding protein sequences is carried out using the Universal Genetic Code. Out of all the possible combinations of 20 amino acids, proteins are formed by the possible combinations that occur naturally. This leaves a large number of unknown combinations of protein sequences that include the Never Born Proteins. A Never Born Protein is a theoretically possible protein that does not occur in nature or may be selected by evolution in future. Objective: In this study, the "GenNBPSeq" online web server is developed to generate Never Born Protein Sequences and to analyze their sequence and structural stability. Methods: The “GenNBPSeq” server is developed based on the Gray Code and Partitioned Gray Code representations of the Universal Genetic Code combined with the novel Toeplitz matrix approach. The sequence and structure analysis is done by various bioinformatics tools for the sample Never Born Protein sequences. Results: The “GenNBPSeq” server is available at http://bioinfo.bdu.ac.in/nbps and the users can generate Never Born Protein sequences and download them in FASTA formats. The Never Born Protein sequences obtained by the above Toeplitz matrix approach contain the same amino acid composition. They also form protein secondary and 3-Dimensional structures with intrinsic stability. Conclusion: This study conjectures that the Never Born Protein Sequences generated by “GenNBPSeq” server using the Toeplitz matrix approach may exhibit intrinsic structural stability. Synthesizing these Never Born Proteins and analyzing their biological applications are major research areas in Systems and Synthetic Biology.

Background: Inferring feature importance is both a promise and challenge in bioinformatics and computational biology. While multiple biological computation methods exist to identify decisive factors of single cell subpopulation, there is a need for a comprehensive toolkit that presents an intuitive and custom view of the feature importance. Objective: We developed a Feature-scML, a scalable and friendly toolkit that allows the users to visualize and reveal decisive factors for single-cell omics analysis. Methods: Feature-scML incorporates the following three main functions: (i) There are seven feature selection algorithms to comprehensively score and rank every feature. (ii) Four machine learning approaches and increment feature selection (IFS) strategy jointly determine the number of selected features. (iii) The Feature-scML supports the visualized feature importance, model performance evaluation, and model interpretation. The source code is available at https://github.com/liameihao/Feature-scML. Results: We systematically compared the performance of seven feature selection algorithms from Feature- scML on two single cell transcriptome datasets. It demonstrates the effectiveness and power of the Feature-scML. Conclusion: Feature-scML is effective for analyzing single-cell RNA omics datasets to automate the machine learning process and customize the visual analysis from the results.

Objectives: Ganoderic acid Me [GA-Me], a major bioactive triterpene extracted from Ganoderma lucidum, is often used to treat immune system diseases caused by viral infections. Although triterpenes have been widely employed in traditional medicine, the comprehensive mechanisms by which GA-Me acts against viral infections have not been reported. Sendai virus [SeV]-infected host cells have been widely employed as an RNA viral model to elucidate the mechanisms of viral infection. Methods: In this study, SeV- and mock-infected [Control] cells were treated with or without 54.3 μM GA-Me. RNA-Seq was performed to identify differentially expressed mRNAs, followed by qRT-PCR validation for selected genes. GO and KEGG analyses were applied to investigate potential mechanisms and critical pathways associated with these genes. Results: GA-Me altered the levels of certain genes’ mRNA, these genes revealed are associated pathways related to immune processes, including antigen processing and presentation in SeV-infected cells. Multiple signaling pathways, such as the mTOR pathway, chemokine signaling pathway, and the p53 pathways, significantly correlate with GA-Me activity against the SeV infection process. qRT-PCR results were consistent with the trend of RNA-Seq findings. Moreover, PPI network analysis identified 20 crucial target proteins, including MTOR, CDKN2A, MDM2, RPL4, RPS6, CREBBP, UBC, UBB, and NEDD8. GA-Me significantly changed transcriptome-wide mRNA profiles of RNA polymerase II/III, protein posttranslational and immune signaling pathways. Conclusion: These results should be further assessed to determine the innate immune response against SeV infection, which might help in elucidating the functions of these genes affected by GA-Me treatment in virus-infected cells, including cells infected with SARS-CoV-2.

Background: CENPA is a rare histone variant that regulates various active centromeres and neocentromeres via diverse signal pathways. However, the expression of CENPA correlated with the prognosis of patients in human pan-cancer is still largely underexplored. Objective: To find the role of CENPA in the prognosis and immunotherapy of cancer patients. Methods: In this study, multiple bioinformatic methods, including the ONCOMINE database, TCGA database, GEPIA database, DAVID database, and TIMER database were integrated to comprehensively investigate the prognosis and immunity of CENPA in pan-cancer. Results: The results showed that CENPA was widely expressed in numerous cancer types, including liver cancer, lung cancer, bladder cancer, and gastric cancer. Meanwhile, the increased CENPA expression was significantly related with poor prognosis in breast cancer, lung cancer, and sarcoma. Additionally, CENPA expression had a positive coefficient for immune cell infiltration, including B cells, CD4+T cells, CD8+T cells, neutrophils, dendritic cells, and macrophages. Furthermore, we screened out TGCT, THCA, and LUSC as the most vital cancers correlated with CENPA expression in the immune microenvironment, according to immune score and stromal score. Notably, 47 common immune checkpoint genes were explored in 33 cancer types based on the coefficients of CENPA expression. In addition, CENPA expression was strongly associated with TMB and MSI in various cancers, like BLCA, BRCA, CESC, and CHOL. Moreover, there was a high correlation between CENPA expression and DNA methylation obtained by calculating relatedness coefficients. Enrichment analysis showed that CENPA might be involved in the progression of cancer through cell cycle-related pathways, p53 signaling pathways, and mismatch repair enrichment pathways. Conclusion: Taken together, our results suggested that CEPNA could be considered a promising predictive biomarker affecting prognosis and immune infiltration in human pan-cancer.

Background: A promoter is a DNA regulatory region typically found upstream of a gene that plays a significant role in gene transcription regulation. Due to their function in transcription initiation, sigma (σ) promoter sequences in bacterial genomes are important. σ70 is among the most notable sigma factors. Therefore, the precise recognition of the σ70 promoter is essential in bioinformatics. Objective: Several methods for predicting σ70 promoters have been developed. However, the performance of these approaches needs to be enhanced. This study proposes a convolutional neural network (CNN) based model iProm70 to predict σ70 promoter sequences from a bacterial genome. Methods: This CNN-based method employs a one-hot encoding scheme to identify promoters. The CNN model comprises three convolution layers, followed by max-pooling and a dropout layer. The architecture tool was trained and tested on a benchmark and an independent dataset. We used four assessment measures to determine the prediction performance. Results: It achieved 96.10% accuracy, and the area under the receiver operating characteristic curve was 0.99. Conclusion: According to the comparative results, iProm70 outperforms the current approaches for defining σ70 promoter. A publicly accessible online web server is created, and it is accessible at the website: http://nsclbio.jbnu.ac.kr/tools/Prom70-CNN/.

Background: Breast cancer is one cancer that develops from breast tissue and one of the major reasons of deaths in of women all over the world. The tumor-infiltrating lymphocytes in tumor immune microenvironment are correlated with the prognosis in breast cancer patients, and play an important role in the occurrence and development of breast cancer. Methods: In this study, by integrating the immune gene expression of 20 breast cancer cohorts from the public dataset, an immune-based prognostic score was established. This immune-based prognostic score was found to be correlated with prognosis, stromal score, tumor purity, three famous immune checkpoints, and immune escape mechanism in breast cancer patients. Results: The clinical application of the prognostic score was verified by the breast cancer patients treated with chemotherapy, and good therapeutic benefit of the prognostic score was obtained. In addition, the XGBoost classifier was used to construct for predicting the high and low prognostic score subtypes, and the predictive results indicated that the XGBoost was suitable to predict these two subtypes in breast cancer patients. Conclusion: Based on these results, we believed that the prognostic score may be used as an effective prognostic marker and may provide great help for chemotherapy treatment of breast cancer patients.

Background: RNA-dependent RNA polymerase (RdRp) contributes to the transcription cycle of the SARS-CoV-2 virus with the possible assistance of nsp-7-8 cofactors. Objective: The study aims to investigate the viral protective effects of complementary drugs in computational approaches that use viral proteins. Methods: For the in silico studies, the identified compounds were subjected to molecular docking with RdRp protein followed by structural and functional analyses, density functional theory (DFT), and molecular dynamics (MD) simulation. The 3D structure of RdRp (6m71 PDB ID) was obtained from the protein databank as a target receptor. After reviewing the literature, 20 complementary and synthetic drugs were selected for docking studies. The top compounds were used for DFT and MD simulation at 200 ns. DFT of the compounds was calculated at B3LYP/6-311G (d, p) based on chemical properties, polarizability, and first-order hyperpolarizability. Results were analyzed using USCF Chimera, Discovery Studio, LigPlot, admetSAR, and mCule. Results: Computational studies confirmed the potent interaction of the complementary drugs forsythiaside A, rhoifolin, and pectolinarin with RdRp. Common potential residues of RdRp (i.e., Thr-556, Tyr- 619, Lys-621, Arg-624, Asn-691, and Asp-760) were observed for all three docking complexes with hydrogen bonding. Docking analysis showed strong key interactions, hydrogen bonding, and binding affinities (-8.4 to −8.5 kcal/mol) for these ligands over the FDA-approved drugs (−7.4 to −7.6 kcal/mol). Docking and simulation studies showed these residues in the binding domains. Conclusion: Significant outcomes of novel molecular interactions in docking, simulation, DFT, and binding domains in the structural and functional analyses of RdRp were observed.

Background: As an essential positively charged RNA modification, N⁷-methylguanosine (m⁷G) has been reported to be associated with multiple diseases including cancers. While transcriptomewide m⁷G sites have been identified by high-throughput sequencing approaches, the disease-associated m⁷G sites are still largely unknown. Therefore, computational methods are urgently needed to predict potential m⁷G-disease associations, which is crucial for understanding the biosynthetic pathways of tumorigenesis at the epi-transcriptome layer. Objective: We hope to develop an effective computational method that can accurately predict the associations between m⁷G sites and diseases, and then to prioritizing candidate m⁷G sites for novel diseases. Methods: In this article, we proposed a Schatten p-norm constrained bounded low-rank subspace recovery (SpBLRSR) method for m⁷G-disease association prediction. An m⁷G-disease block matrix was built to alleviate the sparseness during the association pattern discovery process. By incorporating the lowrank representation (LRR) model and sparse subspace clustering (SSC) model, SpBLRSR was designed to capture both the global and local structures of the association pattern. Results Compared with the benchmark methods, SpBLRSR achieved the best performance in predicting associations between m⁷G sites and disease, and in prioritizing m⁷G sites for novel diseases. Then the robustness of Schatten p-norm in our method was further validated via a noise contamination experiment. Finally, a case study of breast cancer was performed to elucidate the biological meaning of our method. Conclusion: SpBLRSR exploits the disease pathogenesis at the epitranscriptome layer by predicting potential m7A sites for disease.