I recently ran RGI on some gut microbiome samples, and I wanted to compare the prevalence of extended spectrum beta-lactamase (ESBL) genes between some of the groups in my study. This wasn’t listed directly in the RGI output, so I needed to add that info in. I asked around and found 3 solutions: The information is listed in the card.json file that you can access following these download instructions. If you’re not familiar with json files, you can get a “pretty” view of that file by running `python -m json.tool card.json`, or there are some online json viewers. h/t @torsten You can get the in ..read more

Subtitle: a series of sickening facts I’m interested in the role of the gut microbiome in infection, especially in settings with a high burden of infectious diseases. Therefore, I did a literature search on PubMed (full text of the search here). The search returned a total of 47 papers. What proportion do you think were investigating charismatic animals like primates or rhinos, rather than humans? (link to twitter question) The answer is 23% !!!! This answer made me feel physically sick! There’s this exciting new(ish) field of the microbiome. One of the main roles of the microbiome is colonisa ..read more

I’m having a play around with some AI tools for literature searching and I came across the autoresearcher tool (link). Here is a comparison of the output of autoresearcher with ChatGPT4. TL;DR autoresearcher will include literature citations for its output, which is nice because ChatGPT currently cannot do this. However, based on my subjective view of this single result, ChatGPT produces an output that is quite a bit better. One advantage of ChatGPT is that you can ask further follow-up questions, asking for some literature on the subject (up to September 2021), or more details on specific ar ..read more

I was interested in getting a list of countries with a “high burden of infectious diseases”. From googling around, I found this nice image from Our World in Data (I love them! Source, with CSV etc). The underlying data from this comes from the Global Burden of Disease (GBD) initiative at the Institute of Health Metrics and Evaluation. But, they didn’t have it broken down by what GBD call the “level 2” causes, but only by Level 1 causes where they combine communicable, neonatal, maternal and nutritional diseases – this includes: HIV/AIDS and sexually transmitted infections, respiratory infecti ..read more

For some unfathomable reason some journals (e.g. Lancet journals) require you to use floating decimal places instead of normal ones. This is a real pain to do, because you can’t just do a find and replace because that would also replace the full stops. This is an opportunity to use a regex! ChatGPT tells us that “Regular expressions, commonly referred to as regex, are a powerful tool for searching, matching, and manipulating text. They are essentially a pattern-matching language used to identify specific sequences within strings of characters.” (see the bottom of the post for a ChatGPTxplanati ..read more

There have been some recent changes to the conda/mamba ecosystem, see obligatory xkcd. This is my current way of creating a fresh install of conda. Setting libmamba as the conda solver should allow you mamba-like speed, while at the same time letting you just install programs like `conda install`. Install miniconda wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh bash Miniconda3-latest-Linux-x86_64.sh Set libmamba as the default solver – https://www.anaconda.com/blog/a-faster-conda-for-a-growing-community conda update -n base conda conda install -n base conda-lib ..read more

This is a “blog it so I don’t have to ask again on twitter post”. We have a narrative review that we want to post as a pre-print, but Medrxiv/Biorxiv don’t accept them. I asked for pre-print servers that do accept narrative reviews on twitter, and these are the sites that were recommended: SciELO OSF Preprints Preprints.org There’s also a comprehensive list of pre-print servers and their characteristics (scope etc) maintained here. To see whether they accept narrative reviews, you have to add the “content types accepted” column by clicking on the little eye symbol, or download the CSV/Excel an ..read more

TL;DR SKA is a high quality bioinformatics tool which enables you to rapidly do genomic epidemiology of very closely related isolates. SKA shows good correlation with traditional SNP calling techniques (BWA + GATK) SKA shows a consistent trend of identifying around 10 more SNPs between a sub-set of pairs of isolates than traditional method, I’m not sure why as yet (this could be something I’m doing, not the tool itself!) SKA is a tool published a few years ago by Simon Harris. I’ve wanted to give it a go for a while, but haven’t really had any need for it. However, I recently got a nanopore ..read more

This post is written in the spirit of “if you write it in an email to more than one person, turn it into a blog”. It’s for beginners who are just starting out in bioinformatics and who need information about which path to go down for which programming language. ————————— Firstly, you will need to understand how the command line works, because most bioinformatics tools don’t use Graphic User Interfaces, but rather use the command line (aka the terminal). The command line is a powerful way to use a computer, but it has a learning curve, and also there is less of a safety net so you can make mist ..read more

Credit for this post is really due to Peter van Heusden, Torsten Seemann, (and Anders Gonvalces Da Silva has had lots of useful discussions with Torsten about this issue), Conor Meehan and Leonardo de Oliveira Martins. TL:DR Two ways to easily give IQ-TREE the number of constant sites: From Leo – For whole genome phylogenetic trees, you can feed IQ-TREE your whole genome alignment (including constant sites) and it will run in a reasonable time if you pass it the following params `-t PARS -ninit 2`. If you use the IQ-TREE default parsimony options then it doesn’t remove repeated patterns (inc ..read more