Vol 19, No 3 (2024)

Life Sciences

Deep Learning for Clustering Single-cell RNA-seq Data

Zhu Y., Bai L., Ning Z., Fu W., Liu J., Jiang L., Fei S., Gong S., Lu L., Deng M., Yi M.

Abstract

The development of single-cell RNA sequencing (scRNA-seq) technology provides an excellent opportunity to explore cell heterogeneity and diversity. With the growing application of scRNA-seq data, many computational clustering methods have been developed to further uncover cell subgroups, and cell dynamics at the group level. Due to the characteristics of high dimension, high sparsity and high noise of the scRNA-seq data, it is challenging to use traditional clustering methods. Fortunately, deep learning technologies characterize the properties of scRNA-seq data well and provide a new perspective for data analysis. This work reviews the most popular computational clustering methods and tools based on deep learning technologies, involving comparison, data collection, code acquisition, results evaluation, and so on. In general, such a presentation points out some progress and limitations of the existing methods and discusses the challenges and directions for further research, which may give new insight to address a broader range of new challenges in dealing with single-cell sequencing data and downstream analysis.

Current Bioinformatics. 2024;19(3):193-210
pages 193-210 views

Mathematical Modelling and Bioinformatics Analyses of Drug Resistance for Cancer Treatment

Li L., Zhao T., Hu Y., Ren S., Tian T.

Abstract

Cancer is a leading cause of human death worldwide. Drug resistance, mainly caused by gene mutation, is a key obstacle to tumour treatment. Therefore, studying the mechanisms of drug resistance in cancer is extremely valuable for clinical applications.

:This paper aims to review bioinformatics approaches and mathematical models for determining the evolutionary mechanisms of drug resistance and investigating their functions in designing therapy schemes for cancer diseases. We focus on the models with drug resistance based on genetic mutations for cancer therapy and bioinformatics approaches to study drug resistance involving gene co-expression networks and machine learning algorithms.

:We first review mathematical models with single-drug resistance and multidrug resistance. The resistance probability of a drug is different from the order of drug administration in a multidrug resistance model. Then, we discuss bioinformatics methods and machine learning algorithms that are designed to develop gene co-expression networks and explore the functions of gene mutations in drug resistance using multi-omics datasets of cancer cells, which can be used to predict individual drug response and prognostic biomarkers.

:It was found that the resistance probability and expected number of drug-resistant tumour cells increase with the increase in the net reproductive rate of resistant tumour cells. Constrained models, such as logistical growth resistance models, can be used to identify more clinically realistic treatment strategies for cancer therapy. In addition, bioinformatics methods and machine learning algorithms can also lead to the development of effective therapy schemes.

Current Bioinformatics. 2024;19(3):211-221
pages 211-221 views

A Unified Probabilistic Framework for Modeling and Inferring Spatial Transcriptomic Data

Huang Z., Luo S., Zhang Z., Wang Z., Zhou T., Zhang J.

Abstract

Spatial transcriptomics (ST) can provide vital insights into tissue function with the spatial organization of cell types. However, most technologies have limited spatial resolution, i.e., each measured location contains a mixture of cells, which only quantify the average expression level across many cells in the location. Recently developed algorithms show the promise to overcome these challenges by integrating single-cell and spatial data. In this review, we summarize spatial transcriptomic technologies and efforts at cell-type deconvolution. Importantly, we propose a unified probabilistic framework, integrating the details of the ST data generation process and the gene expression process simultaneously for modeling and inferring spatial transcriptomic data.

Current Bioinformatics. 2024;19(3):222-234
pages 222-234 views

A Systematic Review of the Application of Machine Learning in CpG Island (CGI) Detection and Methylation Prediction

Wei R., Zhang L., Zheng H., Xiao M.

Abstract

Background:CpG island (CGI) detection and methylation prediction play important roles in studying the complex mechanisms of CGIs involved in genome regulation. In recent years, machine learning (ML) has been gradually applied to CGI detection and CGI methylation prediction algorithms in order to improve the accuracy of traditional methods. However, there are a few systematic reviews on the application of ML in CGI detection and CGI methylation prediction. Therefore, this systematic review aims to provide an overview of the application of ML in CGI detection and methylation prediction.

Methods:The review was carried out using the PRISMA guideline. The search strategy was applied to articles published on PubMed from 2000 to July 10, 2022. Two independent researchers screened the articles based on the retrieval strategies and identified a total of 54 articles. After that, we developed quality assessment questions to assess study quality and obtained 46 articles that met the eligibility criteria. Based on these articles, we first summarized the applications of ML methods in CGI detection and methylation prediction, and then identified the strengths and limitations of these studies.

Result:Finally, we have discussed the challenges and future research directions.

Conclusion:This systematic review will contribute to the selection of algorithms and the future development of more efficient algorithms for CGI detection and methylation prediction

Current Bioinformatics. 2024;19(3):235-249
pages 235-249 views

Recent Advances in the Phylogenetic Analysis to Study Rumen Microbiome

Wassan J., Wang H., Zheng H.

Abstract

Background:Recent rumen microbiome studies are progressive due to the advent of nextgeneration sequencing technologies, computational models, and gene referencing databases. Rumen metagenomics enables the linking of the genetic structure and composition of the rumen microbial community to the functional role it plays in the ecosystem. Systematic investigations of the rumen microbiome, including its composition in cattle, have revealed the importance of microbiota in rumen functions. Various research studies have identified different types of microbiome species that reside within the rumen and their relationships, leading to a greater understanding of their functional contribution.

Objective:The objective of this scoping review was to highlight the role of the phylogenetic and functional composition of the microbiome in cattle functions. It is driven by a natural assumption that closely related microbial genes/operational taxonomical units (OTUs)/amplicon sequence variants (ASVs) by phylogeny are highly correlated and tend to have similar functional traits.

Methods:PRISMA approach has been used to conduct the current scoping review providing state-ofthe- art studies for a comprehensive understanding of microbial genes’ phylogeny in the rumen microbiome and their functional capacity.

Results:44 studies have been included in the review, which has facilitated phylogenetic advancement in studying important cattle functions and identifying key microbiota. Microbial genes and their interrelations have the potential to accurately predict the phenotypes linked to ruminants, such as feed efficiency, milk production, and high/low methane emissions. In this review, a variety of cattle have been considered, ranging from cows, buffaloes, lambs, Angus Bulls, etc. Also, results from the reviewed literature indicate that metabolic pathways in microbiome genomic groupings result in better carbon channeling, thereby affecting methane production by ruminants.

Conclusion:The mechanistic understanding of the phylogeny of the rumen microbiome could lead to a better understanding of ruminant functions. The composition of the rumen microbiome is crucial for the understanding of dynamics within the rumen environment. The integration of biological domain knowledge with functional gene activity, metabolic pathways, and rumen metabolites could lead to a better understanding of the rumen system.

Current Bioinformatics. 2024;19(3):250-263
pages 250-263 views

Advancements in Yoga Pose Estimation Using Artificial Intelligence: A Survey

Chamola V., Gummana E., Madan A., Rout B., Coelho Rodrigues J.

Abstract

Human pose estimation has been a prevalent field of computer vision and sensing study. In recent years, it has made many advances that have helped humanity in the fields of sports, surveillance, healthcare, etc. Yoga is an ancient science intended to improve physical, mental and spiritual wellbeing. It involves many kinds of asanas or postures that a practitioner can perform. Thus, the benefits of pose estimation can also be used for Yoga to help users assume Yoga postures with better accuracy. The Yoga practitioner can detect their own current posture in real-time, and the pose estimation method can provide them with corrective feedback if they commit mistakes. Yoga pose estimation can also help with remote Yoga instruction by the expert teacher, which can be a boon during a pandemic. This paper reviews various Machine Learning, Artificial Intelligence-enabled techniques available for real-time pose estimation and research pursued recently. We classify them based on the input they use for estimating the individual's pose. We also discuss multiple Yoga posture estimation systems in detail. We discuss the most commonly used keypoint estimation techniques in the existing literature. In addition to this, we discuss the real-time performance of the presented works. The paper further discusses the datasets and evaluation metrics available for pose estimation.

Current Bioinformatics. 2024;19(3):264-280
pages 264-280 views

Application of Improved Support Vector Machine for Pulmonary Syndrome Exposure with Computer Vision Measures

Khadidos A., Alshareef A., Manoharan H., Khadidos A., Selvarajan S.

Abstract

Background:In many medically developed applications, the process of early diagnosis in cases of pulmonary disease does not exist. Many people experience immediate suffering due to the lack of early diagnosis, even after becoming aware of breathing difficulties in daily life. Because of this, identifying such hazardous diseases is crucial, and the suggested solution combines computer vision and communication processing techniques. As computing technology advances, a more sophisticated mechanism is required for decision-making.

Objective:The major objective of the proposed method is to use image processing to demonstrate computer vision-based experimentation for identifying lung illness. In order to characterize all the uncertainties that are present in nodule segments, an improved support vector machine is also integrated into the decision-making process.

Methods:As a result, the suggested method incorporates an Improved Support Vector Machine (ISVM) with a clear correlation between various margins. Additionally, an image processing technique is introduced where all impacted sites are marked at high intensity to detect the presence of pulmonary syndrome. Contrary to other methods, the suggested method divides the image processing methodology into groups, making the loop generation process much simpler.

Results:Five situations are taken into account to demonstrate the effectiveness of the suggested technique, and test results are compared with those from existing models.

Conclusion:The proposed technique with ISVM produces 83 percent of successful results.

Current Bioinformatics. 2024;19(3):281-293
pages 281-293 views