1/5 Our preprint about software crisis in computational biology is live! We have performed a review of usability and archival stability of published omics tools and resources. We have found that 28% of the tools are impossible to install. bit.ly/2yAphZX 
@thmosqueiro just suggested that usually people post threads replying their own tweets, so that twitter group them together :)
I am new to twitter so i just made 5 separate tweets. Let me get them together here.
I am new to twitter so i just made 5 separate tweets. Let me get them together here.
2/5 This is a collaboration with many briliant collegues : @thmosqueiro, @gargidayama @ferocious_sloth @zarlab @jonathan_flint1 @blekhman. And of course, excelelent undegraduates, who has spent 72 hours to install 99 randomly selected tools! bit.ly/2yAphZX
3/5 Amont all the tools only 50% are easy-to-install and 21% are hard-to-insall. Tools warapped as @anacondainc are 100% instalable. Majority of tools fail automatics instalation test! Tools which are possible to install are significantly more cited!
4/5 Among 24,490 URLs of published omics software resources 26% are broken. No effect of the journal impact factor on the availability of published links. Accessible links have more attention in social media. @github now account for 20% of all URLs and has only 1% of broken URLs
5/5 The take-home messages:
1. Host software on archivally stable services e.g. @github
2. Provide easy-to-use installation interface (e.g. @anacondainc ) and get more citations
3. Tools wrapped as @anacondainc are 100% instalable!
Read more here bit.ly/2yAphZX
1. Host software on archivally stable services e.g. @github
2. Provide easy-to-use installation interface (e.g. @anacondainc ) and get more citations
3. Tools wrapped as @anacondainc are 100% instalable!
Read more here bit.ly/2yAphZX
