@BorisLenhard@AMartinezArias@ItaiYanai Arxiv family of platforms is completely sufficient for presenting scientific work to the community. In which journal was the the transformers paper published, or convnets or bowtie or MACS? Nobody really cares. If the work is good people will start using it.
@TeifLab@RalfBlossey@ItaiYanai The ultimate proof of peer recognition is when science is used to help reduce human suffering. You really think Fleming cared about citations?
@StevenSalzberg1@NatureAg Although the mentioned study provides limited findings, there are multiple lines of evidence showing discontinuous nature of aging in humans:
https://t.co/laadgUk5Mp
I wonder if GNU Guix could also be an alternative. currently, there are 28000 packages, and many data science packages are there including PyTorch. https://t.co/U3yEUNcau5
harder to package for GNU Guix as it is built for reproducibility, but they have been building some automation tools lately
We hypothesized that boosting RNAi could boost antiviral immunity. In vivo. We genetically engineered a mouse with 10x stronger RNAi and infected it with CVB3, TBEV, EMCV, LCMV. It didn't work. Mice are available to anyone who wants to test other viruses.
https://t.co/mr55H0AeNJ
@mo_lotfollahi@dagarfield The tools for seamless remote collaboration have been around for the last 10 years (i.e. internet infrastructure), and it's really hard to measure the disruptiveness of recent work (if we assume they even measure such trait).
@NimwegenLab Very interesting! Two questions - did you try running on both exonic and intronic reads? Does it increase the number of states?
Would a metacell approach (separheaded by Tanay lab) be useful for dataset size reduction?
@IsabelleStevant Would be nice to see how many TRs serve as stage specific promoters, and whether they have different TF binding affinities, and to try to deconvolve the germline from the somatic TR activation (can probably be done based on single cell RNAseq expression)
@IsabelleStevant Awesome work! It would be interesting to correlate the activity with the age of the elements - are younger/older elements active. Especially with regards to SINEb2 which have a CTCF motif.
We report that simply clustering a MSA lets #AlphaFold predict multiple states of fold-switching proteins.
We used this to screen for alt states in proteins without known fold-switching, and describe one such predicted novel state for the DsbE enzyme. 1/n
https://t.co/ql6TiZ7r9l
@blekhman@Y_Gilad Along with the lack of standardization, lack of quality, lack of professionalizm, lack of determinism, and most importantly, lack of repercussions for uncollegial behavior, discriminatory language, and factual errors.