Quick update for the nf-core meta-omics pipeline f...
# meta-omics
e
Quick update for the nf-core meta-omics pipeline flow use case: I started playing with some of the datasets in the study that @Daniel Lundin suggested (Brackish surface water), that contain MT, MG and Amplicon samples, which is exactly what we want. Already ran
fetchngs
for the pilot use-case, and currently running
metatdenovo
on our cluster. It doesn't make sense to run
detaxizer
since there is no host in this environment. I made my use-case repo public, so feel free to have a look and make any PRs with suggestions/corrections etc: github.com/vagkaratzas/meta-omics-sig Something that you might find insightful, and gives flavour to the project: vagkaratzas.github.io/meta-omics-sig I made a github page for the repo, that can act as both project tracking resource, as well as educational material, since users can click on a pipeline and see a couple of sentences as info, expected inputs/outputs (chained pipelines), and links to each nf-core site
100 rainbow 2
❤️ 1
🤩 1
d
Nice! 🙂 You should add rRNA removal with
--sequence_filter
. I use Silva's
SILVA_138.1_allrRNAref_NR99_tax.fasta.gz
.
The pipline conveyor is pretty! Phyloplace could well run both on metatdenovo and mag output, possibly with proteinfamilies in between. Let's discuss this in person on occassion.
❤️ 1
e
where is seqsubmit, malaka? 😁
😆 1
e
Sure! The bots suggested that mode I think, but said it is going to be heavy, so going for a first lightweight run to see until where we can reach, and then start adding more biological sense, and more runs :D
Tomorrow @Ekaterina Sakharova..need to update the metaomics map now that it's released
🙌 1
j
@Evangelos Karatzas depending on the question, I disagree that detaxizer doesn't make sense: 1. You could have sample handler/lab scientist contamination, and the human DNA from when they sneezed on a sample could cause false positive hits if you have contaminated reference genomes in some of the pipelines (ok probably the fraction will be very low given these are contemporary samples, but still!) 2. It would be overkill but you could use detaxizer to remove sequencing spike in controls like phiX
👀 1
e
Alright, makes sense!
j
I.e. maybe weak, but not un justifiable, depending on what you are trying to do here
e
we will definitely launch more datasets touching more pipelines (the github pages site is built with that in mind), but I think detaxizer can be run for this one as well..curious to see if it finds any hits