Looking forward to introducing the OSU Microbiome Platform tomorrow! Join in person or virtually to hear how it might support your #microbiome #metagenomics research. @OSUmicrobiome @IDIatOSU https://t.co/MumwiEYRPJ

Go Bucks!!🏈🏈

Dress rehearsal before the big show tomorrow night! #TBDBITL #GoBucks

First week with @IDIatOSU is in the books - didn't manage to get everybody together for a pic (where were you @Dr_SFaith😉), but big thanks for the welcome AMSL!!!

Just wrapped up the MCIC/@CAPS_OSU #bioinformatics @osc workshop - many thanks to @JelmerPoelstra for all his efforts in making this happen, and to everyone who participated! #supercomputing #genomics

In the first study analyzing host & pathogen genes in climate change conditions, @OhioState researchers found that in high heat & low water conditions, infected pine trees starve & disease-causing fungi thrive. @CFAES_OSU

Impactful science supported by @CAPS_OSU from @OSUPlantPath PhD candidate @soumya4evr, CAPS members Enrico Bonello, @fungolution, and @MikeSovicOSU, and others!! #climate #GlobalWarming #pathogens #plants #forests @CFAES_OSU

How can #climatechange #globalwarming conditions affect plant susceptibility to #pathogens? Work led by @soumya4evr & Bonello Lab of @OSUPlantPath is now out – YAY! #plantdisease @CAPS_OSU @IDIatOSU @OhioStSustain https://t.co/r27agq3M6r

Retweeting this to add the #Rstats tag, as I expect there may be folks in that community who might be able to help out here...

Hey #Bioinformatics folks, @Oleks_Zinenko is a colleague in Ukraine🇺🇦looking for US-based mentors (remote) for short-term student projects. Spread the word & contact me (or him) if you can help. Details linked below - apparently bio was added recently. https://t.co/9jZcK1Gls1

(19/19) 3.) Wonder how widespread this issue was. Doubt this is only RNAseq data analyzed by that pipeline. Had this not been 2nd set of samps, with initial results to compare, not sure lab would've questioned “no DEG” result at all & RNAseq part of paper likely wouldn’t exist.

(18/19) 2.) Lack of #ReproducibleResearch & transparency from company led to weeks of hands-on effort & months of project delay trying to uncover the small mistake. All unnecessary - easily ID'd and fixed if sufficient detail was given, or at least provided on request.

(17/19) 1.) An initial small mistake had significant implications for inferences, but hey, mistakes happen. Be careful with those parameter settings. #bioinformatics

(16/19) Takeaways here? At least a few…

(15/19) Seemed to be trying anything to avoid saying 1st analysis was just based on wrong strand, which now we had 💪evidence for. They did slip a line in 1 message that they “went through (our) analysis results and confirmed that (they were) reliable”. So...thx for that I guess?

(14/19) I ?'d the exon vs gene suggestion in a couple ways. Response was to propose another cause for diffs - removal of rRNA before mapping. Nope. More discussion amongst their informatics team and they said a new genome version might have caused discrepancy. Nope again.

(13/19) Instead, data were ignored and rep’s response focused on their 2nd analysis, saying they had forgotten to send me one argument they used – they'd analyzed on exons vs genes, but didn’t include that in the code they sent. This wasn’t making sense on several levels.

(12/19) I sent the results, with detailed explanation, hoping they could just confirm their original analysis was incorrectly based on strand 2 and we could move on. No luck.

(11/19) Tried replicating their analyses with what info I had, guessing at param settings where necessary, and comparing results to their read count tables. Finally got something nearly identical to the their first analysis and as we suspected, it was counting on wrong strand.

(10/19) Still no code, but they were “re-analyzing...with adjusted parameters”. Later got new results w/partial code - 30 DEGs. Still not close to our 200+, nor did the identities align w/prev results. But again, they attributed differences to software used, thresholds, etc.