BOL: Related items

Calling variants in non-diploid systems

Neel — Sat, 26 Jun 2021 15:37:49 -0500

The main challenge associated with non-diploid variant calling is the difficulty in distinguishing between the sequencing noise (abundant in all NGS platforms) and true low frequency variants. Some of the early attempts to do this well have been accomplished on human mitochondrial DNA although the same approaches will work equally good on viral and bacterial genomes (Rebolledo-Jaramillo et al. 2014, Li et al. 2015).

Address of the bookmark: https://training.galaxyproject.org/training-material/topics/variant-analysis/tutorials/non-dip/tutorial.html

GenomeTools: The versatile open source genome analysis software

Neel — Wed, 02 Feb 2022 04:00:21 -0600

The GenomeTools genome analysis system is a free collection of bioinformatics tools (in the realm of genome informatics) combined into a single binary named gt. It is based on a C library named “libgenometools” which consists of several modules.

If you are interested in gene prediction, have a look at GenomeThreader.

http://genometools.org/pub/

Address of the bookmark: http://genometools.org/

CrossMap: program for genome coordinates conversion between different assemblies

Rahul Nayak — Tue, 25 Jan 2022 17:59:32 -0600

CrossMap is a program for genome coordinates conversion between different assemblies (such as hg18 (NCBI36) <=> hg19 (GRCh37)). It supports commonly used file formats including BAM, CRAM, SAM, Wiggle, BigWig, BED, GFF, GTF, MAF VCF, and gVCF.

Address of the bookmark: http://crossmap.sourceforge.net/

Monkeypox virus isolate MPXV_USA_2022_MA001, complete genome

Jit — Tue, 26 Jul 2022 06:21:07 -0500

LOCUS       ON563414              197205 bp    DNA     linear   VRL 30-MAY-2022
DEFINITION  Monkeypox virus isolate MPXV_USA_2022_MA001, complete genome.
ACCESSION   ON563414
VERSION     ON563414.3
KEYWORDS    .
SOURCE      Monkeypox virus (monkeypox)
  ORGANISM  Monkeypox virus
            Viruses; Varidnaviria; Bamfordvirae; Nucleocytoviricota;
            Pokkesviricetes; Chitovirales; Poxviridae; Chordopoxvirinae;
            Orthopoxvirus.

Address of the bookmark: https://www.ncbi.nlm.nih.gov/nuccore/ON563414

bacLIFE: an automated genome mining tool for identification of lifestyle associated genes

BioStar — Fri, 15 Mar 2024 04:59:14 -0500

bacLIFE is a streamlined computational workflow that annotates bacterial genomes and performs large-scale comparative genomics to predict bacterial lifestyles and to pinpoint candidate genes, denominated lifestyle-associated genes (LAGs), and biosynthetic gene clusters associated with each lifestyle detected. This whole process is divided into different modules:

Clustering module Predicts, clusters and annotates the genes of every input genome
Lifestyle prediction Employs a machine learning model to forecast bacterial lifestyle or other specified metadata
Analitical module (Shiny app) Results from the previous modules are embedded in a user-friendly interface for comprehensive and interactive comparative genomics.

You can find the complete wiki here [https://github.com/Carrion-lab/bacLIFE/wiki/bacLIFE-wiki]

Address of the bookmark: https://github.com/Carrion-lab/bacLIFE

Can AI Decide Which Genomic Model to Trust?

BioStar — Sat, 24 Oct 2026 03:10:06 -0500

Imagine you're sitting with a genomic dataset of some sort (thousands of SNPs), a set of plant phenotypes and the one very simple question: can we predict what these plants will become? The first thing you'd probably do is reach for machine learning. That's where the real problem lies. Which model do we use? GBLUP? Bayesian ridge? LASSO? Random Forest? A neural net? How do we know that the model that looks so good on a single cross-validation split will still be good tomorrow? In genomic prediction we're often just left to pick these things manually. The problem is that a single cross-validation split can make a model look incredibly good when in reality it's not. This is the problem that Nagendra Pratap Singh and Venugopal Mendu have set out to solve with GenoBridge (https://github.com/nps-genomics/GenoBridge), a sample size adaptive machine learning pipeline designed to reduce the amount of human guesswork involved. The paper (Plant Methods, September 2026) describes a pipeline that instead of asking the user to pick a model and parameters in advance, adapts the analysis based on the size of the sample. It tests multiple models and evaluates them via cross-validation. The premise is simple but powerful: don't just ask if a model can predict, ask if the model can predict when the data it's being tested on are genuinely unseen. GenoBridge therefore places a strong emphasis on cross-validated R2 as well as correlation, using the reproducibility of the predictive signal as a gate before moving into GWAS. In other words prediction is the first checkpoint: if the genomic signal can't generalise then we shouldn't be trusting the downstream association results. Once a trait has passed that checkpoint it proceeds into kinship based mixed model GWAS to provide a bridge between prediction and biological interpretation.

The authors tested the pipeline across 5 datasets from 4 species, spanning an almost 80-fold difference in sample size (from 136 to over 10,700 individuals) and found that the predictive accuracy was within 0.02-0.05 of the standard GBLUP and Bayesian ridge regression approaches, again without the user having to manually select a model. More importantly the pipeline was able to recover known biological signals (FLC, FT and DOG1 in Arabidopsis, Waxy and GW5 in rice and Glu-D1 in wheat) and produced well calibrated association statistics. The most interesting observation though is that prediction accuracy and biological association are not necessarily the same thing. Two traits with identical predictive performance could have association significance that varied by over 60 orders of magnitude. So knowing how well we can predict a trait doesn't necessarily tell us anything about the distribution of the genetic architecture of that trait across the genome. That's where GenoBridge is actually of value. It's not just another machine learning pipeline. It's an example of an adaptive analytical workflow that doesn't just throw algorithms at the data but asks whether the data actually justify the next analytical step. As the amount of genomic data we generate continues to grow and the number of computational methods available to us increases, the bottleneck may not be in finding the right algorithm at all. The bottleneck may be in deciding which algorithm we can actually trust.

GenoBridge offers a potential solution: let the data tell you which model is appropriate, validate the prediction rigorously and only then move from prediction to biological discovery. That may be where computational genomics is heading. Rather than a future in which there is one single 'best' model, we may be heading towards a future of intelligent systems that know when a model is appropriate, when a result is reliable and when the biology is actually telling us something that we should be listening to.

More at https://link.springer.com/article/10.1186/s13007-026-01596-5

Claus-Peter Stelzer Lab

Mon, 15 Mar 2021 15:24:41 -0500

Interested in various topics at the intersection of ecology and evolution. In my research I use rotifers as model organisms for experimental studies at the individual and population level. Rotifers are ideally suited for this, because populations of thousands can be kept in small containers in the lab, while single individuals can still be handled conveniently.

More at https://www.uibk.ac.at/limno/personnel/stelzer/index.html.en#research

Genome + Epigenome = New Understanding of the Pathogens in Your Food

Fri, 27 Sep 2013 11:30:45 -0500

UC Davis's Bart Weimer describes foodborne pathogens and their proclivity for rapid genome rearrangement. The 100K Pathogen Genome Project he leads is using PacBio long-read sequencing to close genomes and analyze methylation; Weimer reports that his team has already discovered new epigenetic modifications in Salmonella and Listeria with the technology. www.pacb.com/microbe

Bioinformatician duties and jobs !!!

Jitendra Narayan — Wed, 05 Mar 2014 14:32:26 -0600

Needle in a haystack ... ohh yes this is what bioinformatician do. We handle and analyse, Terabytes and Petabytes of genomic data on daily basis.

How to sequence the human genome - Mark J. Kiel

Fri, 30 May 2014 13:24:11 -0500

View full lesson: http://ed.ted.com/lessons/how-to-sequence-the-human-genome-mark-j-kiel Your genome, every human's genome, consists of a unique DNA sequence of A's, T's, C's and G's that tell your cells how to operate. Thanks to technological advances, scientists are now able to know the sequence of letters that makes up an individual genome relatively quickly and inexpensively. Mark J. Kiel takes an in-depth look at the science behind the sequence. Lesson by Mark J. Kiel, animation by Marc Christoforidis.