PhyloStars/Planning And Meetings
November 20, 2013
present: Arlin, Brian Sidlauskas, Joachim Baran, Julie Allen, Mark Westneat, Mike Rosenberg
- update from Arlin
- feedback from 2 people I contacted
- currently working on 3 possible candidates for organizing team
- goal is to get 2 to 3 more people, then we'll reschedule meeting and re-start the process
- met last week with Brian et al to discuss tutorials
- discussion of phylostars plan
- wish list item: reply-by-email (so user's don't have to leave their email client to respond to a thread)
- possibility of other tutorials
- discussion of extra HIP funds
- money is not tied to PhyloStars, can be used for PhyloTastic
- best target is "autogenerate trees for web sites"
- but we can't figure out best strategy to make that happen
- where can we find staff to commit to this for > a few weeks? NESCent?
- possibility that we might need to hold onto this money in case BioSync funds get clawed back
- action items:
- Mark will try to get BioSync funds released to NESCent while opportunity lasts
- Julie will check up on Mesquite list, solicit ideas for tutorials
- Mark will pursue possibility of a tutorial on SNP-to-phylogeny workflow in Python
- Joachim will talk to 2 individuals who added a huge amount of content to biostars in the early years
- Arlin will continue his recruiting efforts
- Arlin, Mike, and Devin will work on tutorials
November 13, 2013
present: Brian, Arlin, Mike Burns (U. Oregon), Devin Bloom (U. Kansas)
We talked about possible tutorial topics.
- Arlin & Mike will work on interpreting phylo trees
- Devin will work on alignment concatenation using GUI tools
- we'll try to finish in a month
November 6, 2013
Attendees: Arlin, Brian Sidlauskas, Joachim
- CIPRES feedback
- they have a (somewhat new) e-mail list, but its not highly active
- discussed BioStars advantages
- questions can be looked up (knowledge base, searchability, editing)
- community can answer questions
- update from Istvan
- implementing sub-domains (doing that for galaxy; in talks with them)
- sub-domains will be able to get a data dump
- sub-domains will be included in the main site (to be confirmed; sub-domains should only be a view)
- Possible people to contact (for us to improve our familiarity with BioStars, and recruit more organizers)
- about their enthusiasm about BioStars - the top 2 historic contributors to BioStars
- Phylo-savvy BioStars users (3 contacted by Arlin)
- WormBase (Joachim)
- use as knowledge base
- questions/answers for WormBase usage, data provenance, curation processes, etc.
- link-out to our tools (web-site, WormMine, WormBase GBrowse)
- use as knowledge base
- summary of BioStars features desired to support specific projects
- address (sub-domain)
- distinctive landing site
- branding (logo, link-outs-- all this could be in a custom header)
- filtered content (see only project-relevant questions)
- get a data dump for the sub-domain
- privileges for “managers” of a sub-domain
- notifications (RSS feed for a sub-domain)
- possible icon for tags (so that tool tags can have their own little branding)
- update & discussion on tutorials ideas (Brian)
- need direction on topics and format
- suggested criteria for a tutorial
- plan for Brian, Arlin and 2 others to meet to choose targets
October 23, 2013
present: Arlin, Julie, Mark, Mike (delayed due to fire alarm)
We reviewed the schedule of preparations for the planned meeting the 2nd week of Feb. We all agreed that the PhyloStars idea is a great idea and needs to happen. But the low number of organizers at the hangout perhaps deepened the sense that our group does not have sufficient energy to carry through the project as planned, on the planned schedule. Therefore we decided to step back and delay the PhyloStars meeting, at least until March or April if not later.
There may be multiple reasons for the lack of energy, but one reason that we pinpointed (and that can be addressed) is uncertainty about what is going to drive the success of the planned meeting. Most of us have some uncertainty because we haven't been involved in BioStars and we don't understand the motivations.
So, to address this issue, those of us who were present today committed ourselves to doing 2 things:
- seek out someone who participates in BioStars (or StackOverflow or SeqAnswers or whatever) and ask them what motivates them. What would motivate them to come to our PhyloStars meeting?
- write down the names of a few people who might bring some more energy to the organizing team.
Once we address the issues of organizer energy and participant motivation, we can get back on track.
As a separate issue, we discussed briefly a target for a separate phylotastic meeting.
October 9, 2013
Notes from telecon, 10/9/2013
some comments and analysis of biostars
- online analysis by Joachim Baran
- hollow curve of months of activity-- drops sharply from 1
- questions and answers. reduction in answers and votes over time
- 2 top contributors played a strong role early in the process
- but BioStars compares favorably to other fora such as seqanswers in terms of features
- want to edit, don't have permissions (Hilmar)
Shannon Oliver's experience with setting up a Q&A forum for iPlant
- motivated by desire to escape burden of ad hoc user support
- required 6 months to feed the effort and build community
Arlin's top 5 from biostars - candidates for tutorials (see Phylostars:_Sample_Questions_and_Answers)
- a few have material for constructing a good tutorial (most don't)
- some require consolidating multiple answers
Julie's interaction with Mesquite list
- got Daisie to talk to principals. They seem interested in trying this.
- discussion of other projects to target
- CIPRES - Arlin can talk to MM
- RAxML - Brian O can talk to Alexis
Questions for Istvan (answered the next day)
- How can I search for *unanswered* questions in a particular topic area? I already figured out that I can choose a tag, then sort by number of answers, then scroll to the bottom of the list-- but I'm looking for something scriptable and elegant.
- this is not supported in the current interface but obviously would be very easy to add, it is just a matter of a creating an SQL query. The feature was not there mainly because this use case has not occurred to anyone yet.
- How stable is BioStars as a project? Is there funding? What should we tell people about how long they can expect BioStars to stay online?
- Well it is an open source project that can be copied/forked etc moreover it is a pure python codebase with no external dependencies, it will run on any system that supports "vanilla" python. All libraries are included in the source. This is mainly to make the code as self sufficient and independent as possible. So anyone that has a data-dump could easily create another Biostar. Now for this same reason we don't publicly distribute the data-dumps, it would be way too easy to have clone sites that peddle ads and worse.
- As for funding yes we do have funding via the a training grant that for a summer workshop on bioinformatics, this will last for another three years - although the funding is more at operational level and supports just a small amount of development. I do that mostly on my own time/cost. But I am planning to submit a small grant that would support a few months of programming effort.
- Is there a way to clean up the tagging? There are separate tags for "phylogenetics" , "tree", "phylogeny", "phylogenetic tree" and "phylogenetic"
- yes moderators can edit tags via a special url (only visible to moderators: http://www.biostars.org/edit/tags/) where tags can be edited, merged and deleted.
- Is there some way to set up email notifications? One of the strengths of an email list is that it tends to maintain users' interest and attention by coming to their inbox every day or every week. We would like for people to have a way to receive customized notifications on a particular topic or set of topics.
- yes this on our current TODO list as next feature - probably be added in the next month or two
- One of the ways that knowledgeable users expect to share their knowledge on BioStars is to improve existing answers. However, new users don't have the status to edit. This is a frustration of some users. What is the pathway that you expect most users to take to getting to the points to edit?
- We tend to promote moderators quite frequently on reputation (this will be made fully automatic, over a certain reputation there will be automatic moderation rights), but if someone would like to moderate all they need to do is ask, especially when a new community would be forming members of the initial group would be directly granted these rights.
- Would there be some way (perhaps depending on some future developments) to create a subdomain with filtered content, e.g., a version of Biostars that shows only the content matching a set of evolution-related tags?
- yes, as the site grows we are thinking about ways to easily filter and reduce content. For example Galaxy team is inquring on using Biostar and their posts will go under a different tab altogether. I really like the subdomain idea though - thanks for brining it up, I am going to make an issue in the tracker - it removes the need for additional tabs and provides a neat way to creates sub-biostars that groups content of a certain type. (https://github.com/ialbert/biostar-central/issues/205)
I would go even further to allow customization both as color and logo of the subdomain so that it is perfectly clear that the user is looking at a subdomain of biostar rather of the main site.
September 25, 2013
- need critical mass of users
- meeting would be set to developing work-flows, tutorials, answering questions from the phylogenetics community. Essentially create momentum and then have the community enhance and grow the content.
- replicate success of R-sig-phylo (but in web interface)
- community engagement strategy (need to engage community before and after meeting)
- identify the audience. who is using biostars to get questions answered, or to convey expertise?
- grad students and postdocs -- always in flux
- involve experts? who can answer questions like in R-sig-phylo
- disseminate questions, examples, etc to stimulate interest
- tutorials? create an example and then engage others to help create tutorials from answers they got from their own questions
- have meeting in February (include remote participants)
- follow-up next summer with conference presentations, booths, etc
- identify the audience. who is using biostars to get questions answered, or to convey expertise?
- how to find stakeholders to participate
- software support for packages? post the question on biostars
- find software packages that do not have mailing lists
- PAML, MrBayes, HyPhy, etc
- Mesquite - talk to people who use Mesquite
- BEAST - tutorials
- how to develop tutorial material
- take a thread from a mailing list and turn it into a tutorial
- add tutorials for the R packages
- ask Joe Felsenstein what are common phylo questions
- how to disseminate our plans so as to energize community to get involved
- put the proposal online
- research gate
- SysBio ask the community what they need.
- ask community for tutorial ideas (evoldir etc)
- go onto biostars and find the 5 biggest questions on Phylogeny
- get a sample of questions use as bait.
- show that there is activity there
- Mark - general advertisement for mailing lists and social media
- Arlin - look for some main questions on BioStars (above)
- Julie - talk to Daisie - get her interested in a basis for answering questions -- what are main questions people ask for Mesquite. What about moving Q&A support to BioStars as a way to increase energy level?
- Brian’s graduate student might be interested?
- Everyone -- write a tutorial? Or ask someone to write a tutorial?
Plans for New Budget
- larger meeting?
- summer students?
- phylotastic code sprint? -- maybe at design stage? -- need full time programmers or grad fellow?
- 3 months of a postdoc?
- undergraduate summer stipend 3 students at NESCent?
- Read through phylotastic proposal -- is there a section a grad student could work on?
- Something broader?
September 11, 2013
- importance of building moment in order to reach critical mass at (or soon after) hackathon
- progress: before-and-after comparison
- need to show folks that phylostars is the fastest way to get a quality, custom answer
- example of r-sig-phylo as success story
- mailing list created from NESCent hackathon
- 960 people
- 100-200 posts per month
- includes people like JF who does not use R but likes to get involved in the theory issues
- importance of having some high-quality tutorials
- 2 pages
- good tutorial can take a day to produce if you are familiar with the topic
- possibility of modifications of BioStars (Istvan)
- availability of history for purposes of before-after comparison