??Halfpipe for RNAseq, ??BioChatter: LLM driven bioscience, ?? CamITree: Phylogenetic Analysis??, PNL: Polygenic Risk Scores using PairNet??
Bioinformer Weekly Roundup
Stay Updated with the Latest in Bioinformatics!
Issue: 74 | Date: 21 February 2025
?? Welcome to the Bioinformer Weekly Roundup!
In this newsletter, we curate and bring you the most captivating stories, developments, and breakthroughs from the world of bioinformatics. Whether you are a seasoned researcher, a student, or simply curious about the intersection of biology and data science, we have got you covered. Subscribe now to stay ahead in the exciting realm of Bioinformatics!
?? Featured Research
This study sequenced and identified lncRNAs and mRNAs expressed in the seminiferous tubules of Guanzhong dairy goats at two developmental stages. By constructing lncRNA-mRNA interaction networks, the researchers identified specific lncRNAs and their target genes that may play crucial roles in spermatogenesis.
This study aims to analyze the differences in gene expression between UC patients and healthy controls using Oxford Nanopore Technology’s long-read RNA sequencing (ONT-RNA-seq). It also explores how alternative polyadenylation (APA) site selection contributes to the pathogenesis of UC.
The study investigates the differential expression of tRNA-derived small RNAs (tsRNAs) in juvenile and adult sheep skin. Using small RNA sequencing and proteomics, researchers identified distinct tsRNA expression patterns between 1-month-old lambs and 24-month-old adult sheep. The analysis revealed 19 highly expressed tsRNAs in juveniles and significant differences in protein expression related to immunity and inflammation.
A fundamental challenge for cancer vaccines is to generate long-lived, functional T cells that are specific to tumor antigens. This article explores how mRNA–lipoplex vaccines targeting somatic mutation-derived neoantigens might address this challenge in pancreatic ductal adenocarcinoma (PDAC), a lethal cancer with few mutations.
The study examined whether cooperative interactions among tumor cells can be used as a therapeutic approach. It identified a mechanism where tumor cells collectively digest extracellular oligopeptides through aminopeptidase secretion, benefiting both secreting and neighboring cells. Disruption of this mechanism, particularly the loss of the key enzyme CNDP2, drives tumor populations to a critical extinction point and reduces tumor growth. This highlights the importance of cooperative nutrient scavenging in tumor survival and reveals a targetable vulnerability.
The study investigates the genetic predisposition of cryptorchidism in rams, focusing on the transcriptome and proteome of cryptic and contralateral testis, as well as normal testis. RNA-seq and TMT-based proteomics identified numerous differentially expressed genes (DEGs) between the testis types. Functional enrichment analysis showed involvement in cell death and metabolic processes for up-regulated DEGs, and reproductive processes for down-regulated DEGs. The study also identified ten candidate key genes related to unilateral cryptorchidism, providing insights into the molecular mechanisms behind spermatogenesis failure.
Genome-wide association study of fat content and fatty acid composition of shea tree (Vitellaria paradoxa C.F. Gaertn subsp. paradoxa) | BMC Genomics
The study conducted a multi-locus genome-wide association study (GWAS) on 122 superior shea trees to identify quantitative trait nucleotides (QTNs) controlling fat content (FC) and fatty acids (FA). It found 47 significant QTNs related to fat traits, with 25 being common QTNs detected by multiple methods. Pathway analysis identified 24 candidate genes involved in FC and FA biosynthesis. The findings enhance the understanding of genetic regulation of FC in shea trees and aid in molecular breeding of high-fat cultivars.
The study extends a synthetic biology-based platform for constructing stable production cell lines to produce recombinant adeno-associated virus (rAAV) serotype 8 (rAAV8). It integrates genetic modules into the HEK293 genome to generate rAAV8 producer cell lines. Productivity was found to be lower than conventional triple plasmid transfection, with Cap protein production and capsid formation as limiting factors. Increasing copies of the AAV8 VP123 gene significantly improved rAAV8 titer. The platform is demonstrated to be robust and applicable to different AAV serotypes.
The study explores a novel RNA regulatory approach using 5-carboxylcytosine (ca5C) base modifications. The technique transforms ca5C into dihydrouracil, causing base mutations that impact RNA functionality. This method effectively manages CRISPR–Cas9 system activities, minimizing off-target effects. The approach advances RNA manipulation and offers a new method for precise gene editing control, with potential applications in chemical biology.
??? Latest Tools
BHCox: Bayesian heredity-constrained Cox proportional hazards models for detecting gene-environment interactions | BMC Bioinformatics The study introduces Bayesian heredity-constrained Cox proportional hazards (BHCox) models to detect gene-environment interactions in survival outcomes. It uses spike-and-slab and regularized horseshoe priors, with the no-U-turn sampler (NUTS) for model fitting. Simulations show the model’s performance, and its application to lung cancer data identifies gene-smoking interactions.
The code is available here.
MFCADTI is a method for predicting drug-target interactions (DTIs) using cross-attention mechanisms to integrate network and sequence features. It evaluates the method on two datasets, showing improved performance over existing methods.
The data and source code are available here.
CamITree is a software for phylogenetic analysis of viral and mitochondrial genomes. It provides a streamlined workflow for gene concatenation, sequence alignment, optimization, and tree estimation using maximum-likelihood and Bayesian methods. CamITree aims to simplify phylogenetic analysis and improve data processing efficiency.
The software and usage manual are available here.?
GCLink is a model for inferring gene regulatory networks (GRNs) from single-cell RNA-sequencing (scRNA-seq) data. It uses graph contrastive learning to aggregate gene features and neighborhood information, improving prediction of gene interactions. Experiments show GCLink outperforms other methods, and pretraining on one cell line with fine-tuning on another demonstrates its effectiveness with limited known interactions.
The source code and data are available here.?
PNL is a tool for optimizing polygenic risk scores (PRS) by integrating models from various PRS methods and using trans-ethnic data. PNL employs the PairNet algorithm, a low-complexity convolutional neural network. Case studies for asthma, type 2 diabetes, and vertigo show PNL matches or improves AUCs compared to other methods.
The source code is available here.
MOSTPLAS is a multi-label learning model designed to predict the host range of plasmids. It uses a pseudo label learning algorithm and self-correction asymmetric loss to handle samples with missing labels. The model was tested on various datasets, demonstrating its ability to recognize more host labels while maintaining high precision.
The source code is available here.
Halfpipe is a computational tool for analyzing RNA-seq data from metabolic labeling experiments to quantify RNA half-lives. It integrates data processing, normalization, and statistical analysis, providing a comprehensive framework for studying RNA stability and dynamics.
The source code is available here.
PhyloFunc introduces a functional beta-diversity metric that incorporates phylogenetic information to analyze metaproteomic data. By leveraging the phylogenetic framework of weighted UniFrac distance, it provides a more informative measure of functional dissimilarities between microbial communities.
The source code is available here.
Cellpose3 is an updated version of the Cellpose tool, designed for improved cellular segmentation in microscopy images. It features a restoration algorithm that enhances image quality, enabling accurate segmentation even in noisy, blurry, or undersampled images. The tool is user-friendly, requiring only a single click for image restoration.
The source code is available here.
TBtypeR is a tool for detecting and classifying mixed-strain infections of Mycobacterium tuberculosis complex. It uses whole genome sequencing data and a phylogenetic SNP panel to identify sublineage mixtures at low frequencies.
The tool is available here.
?? Community News
Biologists have engineered gut bacteria to produce and release proteins within the lower gastrointestinal tract. This method uses bacteriophages to reprogram bacterial cells, potentially overcoming challenges in delivering drugs to this part of the body. The study demonstrated the technique's effectiveness in reducing inflammation and obesity in mice.
Researchers from various institutions, including University of Edinburgh, Vrije Universiteit Amsterdam and Uppsala University, explored the relationship between genetic variants, income, and health. Their study, published in Nature Human Behavior, analyzed genetic data from 668,288 individuals across 12 European countries. The findings suggest that while genetic effects on income and health are small, they are significant and consistent across different populations.
The Chan Zuckerberg Initiative has launched the Billion Cells Project, aiming to create a single-cell dataset of one billion cells. This dataset will be used to train AI models to enhance understanding of cellular behavior and gene function. The project involves collaboration with 10x Genomics, Ultima Genomics, and other researchers, focusing on diverse biological domains such as gene regulation and disease research.
Researchers at Northwestern University used computational structure prediction to design synthetic receptors for cell therapies. They investigated structure-function relationships in engineered receptors, focusing on structural features that impact therapeutic performance. The study highlighted the potential of AI tools like AlphaFold in improving receptor design for better safety and efficacy in cell therapies.
The article discusses various large language models (LLMs) and their applications in research. It highlights how different LLMs excel at specific tasks such as coding, synthesizing information, and generating hypotheses. The guide includes insights from researchers on their preferred AI tools and the strengths of models like OpenAI’s ChatGPT and o3-mini.
Researchers at the University of Arizona College of Medicine – Tucson found that an FDA-approved osteoporosis drug might counter a rare genetic mutation causing dilated cardiomyopathy (DCM). The study used supercomputers and AI to screen 2,000 FDA-approved drugs, identifying one that could potentially correct the structural defect in the heart’s motor protein.
Researchers at the University Hospital Bonn and FAU Erlangen-Nuremberg identified the membrane protein MLC1 as a potential target antigen in multiple sclerosis (MS). Using a combination of B-cell stimulation and protein microarray techniques, they found increased antibody responses against MLC1 in MS patients. The study suggests that MLC1, expressed on astrocytes and neurons, could be a significant factor in the autoimmune response characteristic of MS.
Researchers from the PRAEGNANT Network in Germany explored the clinical implications of ctDNA testing in advanced or metastatic breast cancer. The study involved 49 patients and revealed that 76% had at least one somatic mutation, with common mutations in genes like TP53, PIK3CA, FGFR1, and ATM. The findings suggest that ctDNA testing can provide real-time monitoring and comprehensive genetic profiling, influencing treatment decisions in 35% of cases.
Researchers from Pusan National University have developed a drug delivery strategy using colorectal cancer cell-activated nanoconjugates (CTNCs) inside an alginate matrix. This method ensures highly specific and localized drug release, enhancing the therapeutic efficiency of colorectal cancer treatment.
Bioinformatics is being incorporated into TB surveillance to provide new insights into tuberculosis. Techniques such as genomic analysis, spatial statistics, and machine learning are being used to analyse Mycobacterium tuberculosis genomes, improving TB control and surveillance.
?? Upcoming Events
This webinar will discuss how Large Language Models (LLMs) can accelerate protein classification by automatically generating descriptive annotations, streamlining the traditionally manual curation process. It is designed for bioinformaticians, computational biologists, data scientists, and researchers interested in applying AI to biological problems. The webinar will be held on March 19, 2025.?
This webinar will explore the LLM stack, focusing on open-source frameworks that facilitate deployment at various levels, and discuss the application of LLMs in bioinformatics. It is designed for bioinformaticians, computational biologists, data scientists, and researchers interested in applying AI to biological problems. The webinar will be held on March 26, 2025.?
The webinar "Advancing Biomarker Discovery: Harnessing Multiplex Technology for Disease Diagnostics" aims to provide new insights into disease biomarkers involved in renal dysfunction, long-COVID, and sepsis. It evaluates the potential of multiplex technologies to revolutionize diagnostic workflows and improve patient outcomes1. The event will be held on February 26, 2025.
This webinar will introduce proteomics data in?UniProt?and highlight resources useful for analyzing large protein datasets, including the proteomes service and peptide search. It will also cover programmatic access to the?UniProt?Proteins API and provide coding examples via a Google?Colab?workbook. The webinar will be held on March 6, 2025.?
?? Educational Corner
The article introduces TistheMachineLearner, a lightweight interface designed to simplify the use of Scikit-Learn for classification and regression tasks in both Python and R. It details the functionalities of the two main classes, Classifier and Regressor, and provides examples of their application in machine learning workflows.
This article explains how to create tutorial worksheets for R, Python, Stata, and Julia using Quarto. It covers the process of generating worksheets with and without solutions, highlighting the flexibility and ease of use that Quarto offers for educational purposes.
The article discusses the implementation of one-way repeated measures ANOVA in R. It outlines the steps involved in conducting this statistical analysis, including data preparation, model fitting, and interpretation of results, providing a comprehensive guide for researchers.
?? Connect with Us
Stay connected and engage with us on social media for daily updates, discussions, and more!
?? Subscribe
Don't miss an issue! Subscribe to the Bioinformer Weekly Roundup and receive the latest insights directly in your inbox.
We hope you enjoyed this week's edition of the Bioinformer Weekly Roundup. Feel free to share it with your colleagues and friends who share your passion for bioinformatics!
Disclaimer: The information provided in this newsletter is for educational and informational purposes only and does not constitute professional advice.
Contact: [email protected]
Copyright ? 2025, Bioinformer Weekly Roundup. All rights reserved.