Spice up your bioinformatics skill set with AI ??, Single-cell methods for DNA methylation ??, Quick tips for building FAIR workflows ???
Bioinformer Weekly Roundup
Stay Updated with the Latest in Bioinformatics!
Issue: 11 | Date: 12 October 2023
?? 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're a seasoned researcher, a student, or simply curious about the intersection of biology and data science, we've got you covered. Subscribe now to stay ahead in the exciting realm of bioinformatics!
?? Featured Research
Researchers have created an integrated human neural organoid cell atlas (HNOCA) by compiling data from 36 single-cell transcriptomics sets, offering insight into neural cell types generated in vitro. This atlas, which includes over 1.7 million cells, aids in comparing organoid neural cells to their primary counterparts in the human brain, helping assess organoid fidelity, study disease states, and enhance protocol development.
?? The first author, Zhisong He, has posted a Twitter thread that lays out the main findings of the research.
This study shows that the Infinium BeadChip, a commonly used DNA methylation assay, can be adapted for samples with minimal DNA, including single cells, using both experimental modifications and computational techniques. The refined method provided insights into DNA methylation dynamics in rare mouse primordial germ cells, revealing nuanced chromatin states and aspects of germ cell development.
PRESNER is a name entity recognition (NER) and classification system designed for electronic health record (EHR) prescription data. Utilizing the pre-trained Bio-ClinicalBERT model and trained on UK Biobank data, it excels in extracting medication-related information and categorizing drugs.
?? The corresponding author, Ewan Birney, posted a Twitter thread that explains the motivation behind the work.
Central nervous system tumours pose significant treatment challenges, especially in children. To improve the accuracy of tumour type identification during surgery, a neural network named Sturgeon was developed. Sturgeon uses rapid nanopore sequencing to determine sparse methylation profiles and provides molecular subclassification of these tumours.
Genomic variant benchmark datasets play a pivotal role in advancing genomics and bioinformatics by offering insights into the efficacy of sequencing and analytical techniques. This review delves into existing benchmark datasets and their significance, with special attention to the latest benchmarks concerning medically relevant genes and those with intricate genomic structures.
This study introduces SNAIL, a novel normalization method derived from smooth quantile normalization tailored for co-expression measurements. By using SNAIL, researchers can prevent false-positive associations in co-expression and downstream network analyses, offering significant benefits for large-scale heterogeneous data applications.
This study presents CellPLM, a novel pre-trained model for single-cell data, which innovatively treats cells as tokens and tissues as sentences, addressing the unique challenges of single-cell data. By incorporating spatially-resolved transcriptomic data and a Gaussian mixture prior distribution, CellPLM captures intricate cell-cell relationships and performs better than current models, offering significantly faster inference speeds.
The study introduces MIC-SQTL, a novel algorithm and tool for deconvoluting bulk proteome data by leveraging the relationship between transcriptome and proteome. This tool detects potential cell marker proteins, quantifies cell abundance in mixed proteomes without a reference signature matrix, and facilitates cell-type-specific protein Quantitative Trait Loci (cspQTL) mapping.
The study introduces scTab, an advanced deep learning model tailored for single-cell RNA-seq data, aiming to predict cell annotations across diverse tissues and biological contexts. Unlike traditional models, scTab is designed to handle vast datasets (with training on 22.2 million human cells) and ensures better generalization across different human tissues.
??? Latest Tools
The study introduces FLOP, a tool that assesses the impact of transcriptomics preprocessing choices on downstream gene set enrichment. Using various datasets, it was found that different preprocessing methods can significantly alter enrichment results. FLOP aids in ensuring more reliable biological interpretations from transcriptomic data.
?? The corresponding author, Julio Saez-Rodriguez, posted a Twitter thread that goes into the technical details of the FLOP tool.
The study introduces scPoli, a tool designed for the integration of population-level single-cell atlases, bridging sample metadata with cellular data from diverse sources. scPoli employs generative models for data amalgamation, label transfer, and reference mapping.
?? The corresponding author, Mohammad Lotfollahi, posted a Twitter thread that explains the motivation behind the work.
领英推荐
The study introduces CRAWDAD, a tool designed to quantify spatial relationships between cell types across different scales using spatially resolved omics data. This tool's capabilities are demonstrated on simulated data and real tissues, including the mouse brain and embryo.
?? The corresponding author, Jean Fan, posted a Twitter thread that goes into the technical details of the CRAWDAD tool.
The study introduces BuDDI, a tool designed to bridge the gap between deep cellular profiles obtained from single-cell experiments and bulk profiles from various conditions. BuDDI leverages domain adaptation techniques and a variational autoencoder to simulate cell-type-specific responses to perturbations, factoring in multiple sources of variability.
?? The first author, Natalie Davidson, posted a Twitter thread that provides additional background to the study.
The study introduces simpleaf, a program designed to streamline the use of the alevin-fry ecosystem for single-cell data processing. Simpleaf simplifies workflows, adds new functionalities, and maintains the performance of the original tools. It's available on GitHub, with documentation and tutorials provided.
The study introduces SONATA, a novel computational method designed to address the challenge of integrating diverse single-cell data. Unlike existing approaches that often produce arbitrary alignment solutions, SONATA employs a unique cell-cell ambiguity measurement to gauge alignment ambiguities and robustly enumerates alignment options.
?? Community News
The article explores how incorporating AI and machine learning tools into bioinformatics skill set can enhance data analysis efficiency, leading to faster discoveries and saving researchers considerable time.
This article showcases ?? Ming "Tommy" Tang 's journey from a wet-lab biologist to a computational biologist, emphasising the importance of embracing the command line and learning bioinformatics to handle large data sets.
?? Educational Corner
The newest episode of the Nextflow Channels podcast is now live! Dive into the discussion on computing resources as Ben Sherman talks through the latest and forthcoming runtime enhancements in Nextflow.
This article presents 10 practical tips for workflow developers to apply FAIR4RS principles to their data workflows. Organized by the FAIR acronym, these tips offer a roadmap towards ensuring workflows are more reproducible, sustainable, and aligned with the open science and FAIR standards, ultimately enhancing the quality and longevity of scientific research.
GitHub is set to launch a certification program in early 2024. To join the waitlist, please visit the provided link above.
?? 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.
Editor: James Ashmore | Contact: [email protected]
Copyright ? 2023, Bioinformer Weekly Roundup. All rights reserved.