Congratulations 👏 🎉 to Felicia Wolters to compete two studies at the same time! 🙌
They describe a FAIR compliant plant metabolomics dataset and a systematic profiling thereof!
Thanks to all coauthors!!
#ProudPI in #CompMetabolomics
#metabolomics
See for details:
www.linkedin.com/posts/felici...
PARAS (and its sister algorithm PARASECT) are available at paras.bioinformatics.nl and via antiSMASH.
And above all kudos to #BarbaraTerlouw for her perseverance during multiple years of hard work, and to #ChuanHuang, #DavidMeijer, @jcedielbecerra.bsky.social and #RuolinHe for major contributions!
Thanks to great collaborations with #GregChallis, @wildtypemc.bsky.social , @gillesvanwezel.bsky.social #SerinaRobinson #MattJenner #FabrizioAlberti and many others.
To showcase its practical use, PARAS was employed in collaboration with the Challis group for the experimental characterization of the tryptopeptin biosynthetic gene cluster from Streptomyces strains. 5/n
Benchmarks on independent hold-out data showed that PARAS outperformed previous algorithms. Comparison of a wide range of both older and recent algorithms indicates that the volume of training data, not model choice, is the key determining factor of accuracy in these data regimes. 4/n
Computational analysis of the data also highlighted how different clades of independently evolved A domains can bind the side chains of large substrates using
architecturally distinct pockets, a crucial insight to improve prediction. 3/n
Key to the improvements in accuracy where increases in volume and accuracy of the training data, partly fueled by large-scale community annotations (MIBiG) and followed by painstaking further manual curation to correct errors. For certain taxa, like fungi, training data increased >5-fold. 2/n
Now out in @acs.org JACS Au, the manuscript by #BarbaraTerlouw et al. describing PARAS, a high-accuracy machine-learning algorithm to predict substrate specificities of nonribosomal peptide synthetase (NRPS) adenylation domains, key for estimating natural product structures from BGC sequence. 1/n
Check out this blog from @catarinacarolina.bsky.social on the story behind the paper: communities.springernature.com/posts/making... A story of overcoming challenges, careful engineering and collaborative teamwork.
Our new review article "Microbial secondary metabolites: advancements to accelerate discovery towards application" in Nature Reviews Microbiology is now published!
rdcu.be/d6BHX
thanks @acritschristoph.bsky.social ! And of course we were fortunate that we could build upon some of your excellent earlier work: genome.cshlp.org/content/31/2...
Kudos to @zachreitz.bsky.social for the great work and thanks to funders and collaborators! @erc.europa.eu
Last but not least, it allowed Bita Pourmohsenin from @nadineziemert.bsky.social 's group to reconstruct the evolutionary history of NRP metallophores, which supports that some chelating groups may predate the Great Oxygenation Event.
Several metallophores were then experimentally validated in the lab of Alison Butler at UCSB.
The new method allowed Zach to perform a systematic analysis of metallophore biosynthetic diversity across bacterial genomes, identifying BGCs encoding diverse combinations of chelating moieties.
The approach has a very high precision, and only occasionally fails when these subclusters are encoded elsewhere in the genome. Combining it with metallophore transporter detection further boosts performance.
In this work, @zachreitz.bsky.social developed a systematic approach to predict which BGCs encode production of metallophores, based on carefully tuned rule-based detection of chelating group biosynthetic subclusters.
Are you interested in how to predict functions of natural product biosynthetic gene clusters (BGCs) and their products? And/or do you love metallophores and would love to identify their producers in microbiomes? Check out @zachreitz.bsky.social 's new paper!
elifesciences.org/articles/109... 1/n
Sounds like a great opportunity for a professor position in computational biology in the Netherlands careers.universiteitleiden.nl/job/Assistan...
All in all, thanks to the engineering efforts of Arjan Draisma, @catarinacarolina.bsky.social, Nico Louwen, Satria Kautsar, @jorgenavarro.bsky.social and the rest of the author team, and shout-out to the
@jgi.doe.gov
and
@nigelmouncey.bsky.social .bsky.social
for a wonderful collaboration.
At the same time, for mid-size datasets (most frequent use case), BiG-SCAPE v2 approaches the speed of BiG-SLiCE. For larger datasets, BiG-SCAPE has become more resource-efficient, which now allowed us to analyse >250k BGCs (the antiSMASH database) with it in just a few days on a compute server.
Benchmarking on several different types of ground truth datasets shows the increases in accuracy, with BiG-SLiCE v2 often matching BiG-SCAPE v1 accuracy, and BiG-SCAPE v2 combining its speed increase with a small but notable accuracy gain as well.
BiG-SCAPE was completely rewritten from scratch over the past few years, and now features new alignment modes, a more prominent query mode (which allows you to search genomes from a single BGC query) and more scalable interactive visualizations and data selection/filtering.
Now out in @natcomms.nature.com :
versions 2.0 of both BiG-SCAPE and BiG-SLiCE! With significant speed and accuracy increases, as well as new interactive functionalities.
Read the full paper here #openaccess:
www.nature.com/articles/s41...
Make sure to join us in the MIBiG Annotathons! The MITE database (mite.bioinformatics.nl) will join the efforts! If you are interested in tailoring enzymes/maturases, make sure to join us!
Come join us again in a next round of this massive online open science community effort! 💪
Sign up using the link in the thread.
It’s great fun, and really helps the scientific community. What more can you ask? 🙂
The MIBiG 5.0 Annotathon is coming soon, and registration is now open!
🧬 Does your research involve biosynthetic gene clusters? Do you love natural product biosynthesis? Do you have an interest in rare & exotic enzymes? We can use your help & expertise.
Register here 👉 forms.gle/C1cWcLHtrjT2...