BDB-Lab Summer 2022 Update
Summer Focus: BDB-Lab tools
Our group works on different types of projects, one of them being tool development. This means we produce software that will be used by others. One of the characteristics of tool development is that the projects can be very long-lived and do not end once the paper is published (sustained development being a predictor of software quality):
Some of our tool projects started almost 15 years ago and are still going. Jug’s first commit was in December 2008; the paper came out in 2017, it had a major update in May of this year with important new features. It’s a project in active development.
In the case of mahotas (first commit in August 2008), we are currently in maintenance mode, meaning that bugs are fixed, and packages are updated to newer versions of Python, but we are not adding new features. However, tools are never truly done: just this week, we released a new version of mahotas because there was one function that did not work well with newer versions of NumPy.
Some projects were mostly used internally before having a publication associated with them: Jug was a typical “scratch an itch” tool and was used in several publications before getting its own. Similarly, some of the methods in mahotas got their own publications (e.g., Coelho et al., 2010 and Coelho et al., 2013) before the overall library got its manuscript. In other cases, such as macrel and SemiBin, we intended to write up a manuscript from the start of the project; and you can see in the timeline above that the gap between releasing the software and publishing a manuscript is much narrower.
In the embryonic phase, we have a little tool, called argNorm that maps antibiotic resistance genes (ARGs) to the Antibiotic Resistance Ontology (ARO), so that annotations from different tools can be used together. This will serve as backing to a publication (hopefully preprinted later this year).
Research software quality. We aim for high-quality software that can be counted on by others. Therefore, we make a public commitment to the following six principles:
Five-year support (from date of publication)
Standard, easy to install, packages
High-quality code with continuous integration
Complete documentation
Work well, fail well
Open source, open communication
LOOKING FORWARD
June 30. Svetlana will give a talk on challenges in the metagenomic annotation of antibiotic resistome at the main session Antimicrobial Resistance - Challenges and Perspectives of the #FCOM22.
Aug 14-19. Marija will present the poster “A global survey of the contribution of ecological factors to rates of horizontal gene transfer” at the ISME18 in Lausanne (Switzerland).
Sep 13-15. Anna will give a talk “Exploring the pet gut microbiome within a large-scale investigation of animal gut metagenomes” at the Applied Hologenomics conference in Bilbao (Spain).
Sep 22-27. Svetlana and Luis will attend in September the EDAR6 conference and annual EMBARK meeting in Gothenburg (Sweden).
Oct 15. Svetlana will give the Serbian Society for Bioinformatics and Computational Biology (BIRBI) seminar “Exploring the global resistome within the EMBARK project”.
MANUSCRIPTS
The manuscript describing SemiBin has finally been published: Pan et al. A deep siamese neural network improves metagenome-assembled genomes in microbiome datasets across different environments. Nature Communication 13, 2326 (2022).
CONFERENCES &C
In April, Svetlana presented “Challenges in the metagenomic annotation of antibiotic resistome” at the remote #AMR22 conference and gave an invited lecture on global resistome (University of Novi Sad, Serbia) “Exploring the global resistome within the EMBARK project”.
In May and June, Svetlana gave workshops Intro to bioinformatics at University of Novi Sad (Serbia) to the MSc and PhD students of Microbiology.
Luis presented at the Keystone Symposium on microproteins at the NNF symposium in Copenhagen.
Tutorial. Semibin tutorial happened on June 14th and had 27 registered participants.
Blogpost. Svetlana wrote a blogpost about the nudging heuristic with implausible results by one of the widely used antibiotic resistance gene annotation tools.
Online resources & tools. SemiBin 1.0 was released to coincide with the publication of the manuscript, and a bugfix version (1.0.1) is already out. There was an effort to update several tools:
NGLess v1.4.0 released (many small improvements)
Jug 2.2.1 released. Adds jug pack command
imread updated (some security fixes in the wheels)
The first release of argNorm
PEOPLE
Chengkai Zhu joined the BDB-Lab in June as a postdoctoral researcher. His work currently focuses on the human gut microbiome, such as rheumatic diseases, ankylosing spondylitis, etc. Chengkai is interested in the interaction between microbiome and host and how the microbiome affects human health and shapes the immune system.
Hui Chong will leave the BDB-Lab at the end of June and will join the Beijing Research Center of Huawei in July as a Software Engineer. Best of luck to Hui!
Jelena Somborski joined the BDB-Lab in March as an intern working on big data visualization.
 | A guest post by
|