Nikos Papadopoulos

It took... 📆15 months ⚡️lots of computational time (and electricity😓) 🧐2.5 rounds of review 💾92 code commits 📝260 manuscript commits 🪧3 new DOIs 🧔153 new white hairs on my beard but MorF is now accepted at Genome Biology! genomebiology.biomedcentral.com/articles/10.11… what's new? ⏬ 1/

You might ask: why not submit to a journal? We have too many talented scientists, and too precious little funding, to tolerate the current academic standard of publish-or-perish. We should build an impact-centric nonprofit research culture where we instead publish-and-persist.

@ilyakorsunsky trying to figure out the calculation of E_kb for Harmony - could it be that for Pr(k) I'm supposed to scale R_ki by N rather than N_b? The final formula and python code suggest so. Apologies if you've clarified elsewhere and I missed it!

@thibaut_brunet @zolotarg @epcrocha @mkrupovic @QuintanaMurci @R3RT0 @BIUPasteur @AudeBer @NRascovan @CamLBerthelot @institutpasteur absolutely blew my mind hahaha

@darencard just wanted to say thanks for the extract-fastq-from-bam tutorial. It really saved my mental state this week, and I learned something new as well!

We are very excited to announce the Platynereis Satellite Meeting at the EuroEvoDevo in Finland. Abstracts can be submitted via the meeting registration. We will prioritize Platynereis abstracts but other annelid friends are welcome to submit and attend. @EED2024 #WormWednesday

@YuyaoSong123 @NatureComms Congrats Yuyao! Very needed contribution as cross-species comparisons become more and more accessible 📈

I am exited to share that my first-author paper is online on @NatureComms ! 👩‍🔬 🐭 🐸 🐟 🐵 We compared strategies for cross-species integration of scRNA-seq data, offered guidelines for effective integration and ideas for future method development nature.com/articles/s4146… 🔽more

Thanks to funding by @FWF_at we will put worm snails in the limelight of evodevo and omics research! 2 fully funded PhD positions being announced soon. Feel free to spread the word and send inquiries directly to me if interested. 🐌🪱

@NicolasSMRober1 where would we be without conda envs :O

@galicae Do you *really* need*a conda env? :p. Alternative would be to find a docker image with an r and python env already set up x). For the latest rStudio, I'd encourage to try it out, I quite like it. Even though Rmarkdown already supported python chunks, it had an unfortunate name🙃

every once in a while I try to create another conda environment that has python/R interoperability and I am reminded again that we don't live in the nice timeline

@NicolasSMRober1 I need rpy2 to work, and that is somehow always finicky from inside conda envs D: Maybe trying the R -> python direction works better? I've heard good things about the new Rstudio...

@galicae Reticulate doesn't fit your bill?

@vitaliikl @tangming2005 Yep, that is what I'm suggesting, more or less

@galicae @tangming2005 I am not sure what you mean. Imputation in either direction is using simulated data. It is better to have as direct model as possible. I would attempt to find the same cell clusters or the same txn neighborhoods/metacells in both technologies, then analyze at that resolution.

What tools to use to impute missing genes in spatial data using scRNAseq from the same sample?

@tangming2005 @vitaliikl Wouldn't you be better off mapping the spatial data to normal scRNA and then seeing what expression looks like there?

@vitaliikl We have a gene of interest that is not in the probe list of the spatial data

today I learned that if 1) a gene is expressed a little bit all over the place 2) I am writing a paper/want to sound pompous the correct phrasing is "it is expressed in a punctate pattern" #sciencespeak

May I present to the world: Hirosella fruticosa. A #Bryozoa once known as Plumatella fruticosa has long been ambiguously discussed. We confirm recent phylogen. considerations by #ahmedjoudahnass and erect a new family for this species onlinelibrary.wiley.com/doi/full/10.10… #phdlife #morphology

@bibermair Finally out! Good job, Julian!

@Jeff_Mold Preach 🙏🙏

I need to preface this by saying the data seems sound and this had no real implications on the science described in the paper, but I still had to chuckle at this figure (anonymised slightly)

Are you interested in... - winding scientific journeys🤷‍♀️ - revisiting conventional wisdom🧙‍♂️ - cutting-edge experiments🪚 - combining multiple data modalities🧬🔬⚡️🎥 - a result that can be summarised in one sentence? ...well then, @FabianRuperti has just the thing for you!

this is the equivalent of assembling a genome and then only uploading the raw fastq with instructions on how to assemble🤷‍♀️ Like, sure, that is *technically* reproducible, yay us

this, again🥲Digging through notebooks to piece together the cluster assignments or figure out which cells made it through filtering is not my idea of "reproducible". For the love of God, please upload your final .h5ad/.RDS somewhere🙏