Looking forward
A lot of travel coming up for BDB-Lab members:
Jun 14: Anna will present at the Oxford Nanopore event within ASM Microbe in Atlanta, USA
Jun 17-20: Svetlana will present at the Belgrade Bioinformatics Conference in Belgrade, Serbia.
Jul 1: Anna will present at the Applied Hologenomics Conference in Copenhagen, Denmark
Jul 1: Yiqian will present a poster at the Australian Society for Microbiology in Brisbane, Australia.
Jul 4: Luis will also present at the Australian Society for Microbiology in Brisbane, Australia
Jul 12-18: Luis will be at ISMB2024, followed by the Quest for Orthologs meeting in Montreal, Canada
Jul 18-22: Luis will be in San Francisco meeting collaborators, but has some time to meet anyone who wants to talk science
Aug 18-23: Anna will attend and present a poster at the ISME19 conference in Cape Town, South Africa
Feel free to contact Yiqian, Anna, or Luis if you will be at any of these events!
Focus of the Quarter: AMPSphere
The AMPSphere was first featured in the newsletter in June 2021. At the time, the project was about two years old. This month, the paper describing the resource was published in Cell:
Célio Dias Santos-Júnior*, Marcelo D.T. Torres*, Yiqian Duan, Álvaro Rodríguez del Río, Thomas S.B. Schmidt, Hui Chong, Anthony Fullam, Michael Kuhn, Chengkai Zhu, Amy Houseman, Jelena Somborski, Anna Vines, Xing-Ming Zhao, Peer Bork, Jaime Huerta-Cepas, Cesar de la Fuente-Nunez⁺, and Luis Pedro Coelho⁺ Discovery of antimicrobial peptides in the global microbiome with machine learning in Cell (2024)
It is interesting to read back what we wrote in 2021. At the time, we had the initial database, but no lab data and only preliminary analyses. For example, in that issue of the newsletter, we also welcomed Anna Vines as a remote intern. A few months later, she would run the analyses that are now shown in Fig. 3A of the manuscript!
The AMPSphere is a database of antimicrobial peptides (AMPs) from the global microbiome. An AMP is a small sequence (average size in AMPSphere is 37 residues, with std. dev. of 8). It contains nearly 1 million sequences from >60k metagenomes (which are a part of SPIRE). Importantly, in vitro validation showed that 79/100 are active against at least one strain of bacteria tested. Furthermore, a few are comparable to polymyxin B, a commercial antibiotic. Unlike conventional antibiotics, AMP effects are often strain-specific: a peptide can target a specific strain and not another!1
Most of the candidate AMPs in our database are novel: they do not match any sequence in previously published databases. We also can find some evolutionary origin stories for some of the peptides in the database: they fit a model where small mutations in longer proteins lead to novel peptides. However, for the majority of the sequences, we cannot find any matches, leading to the tantalising possibility that some of them may have been de novo proteins.
This project was the result of a wonderful collaboration with the Cesar de la Fuente Lab at UPenn (with a particular shout out to Marcelo Torres, co-first author) and others around the world. We were lucky that this project also received a decent amount of popular press, with mentions in the BBC (Science in Action), Deutsche Welle, Guardian, El País, Folha de São Paulo, Nature, and many others. Work on analysing this dataset to understand the effect of AMPs on microbes better is ongoing (there are funded PhD positions to do so!).
People.
Congratulations to Shaojun, who successfully defended his PhD!
JP visited Brisbane from South Africa to continue his work on antimicrobial-resistant genes (AMR) in wastewater
Sebastian Dall is visiting Brisbane from Denmark and is working on exploring linkages between nanomotif and SemiBin2
Manuscripts.
The AMPSphere manuscript was published in Cell (see above)
Conferences/talks.
Yiqian presented her work “A catalogue of small proteins from the global microbiome” at the MVIF conference last April, you can see her talk here. You can also read more detail on her work on the preprint
Software tools.
argNorm new releases (versions 0.3.0 and 0.4.0), which contain improved annotations of Resfinder, ARG-ANNOT, and MEGARES
Other.
Check out our new video tutorial on SemiBin2.
Season 2 of the Extremely Open Science project started a few weeks ago, working on “Can we estimate the quality of a genome from an overabundance of pseudogenes?”
Luis will teach a course on Environmental Metagenomics with Physalia. Registrations are open.
Trying to understand this better is ongoing work