Talk:HUB5

From HUB
Jump to navigation Jump to search

Post-HUB5 feedback

Thanks, Matt, for organising another successful HUB!

I thought the two presentations, from Damien and Maik were just right (length, content) for the meeting.

We could have done with more space for the group work, and to just generally have chatted with more people, but I still think it was the right decision to go for the non-Uni venue option, so that there wasn't too much space/time between the meetings, I think HUB thrives on the momentum of fairly frequent meetings. So, wasn't the ideal venue, but at the same time, I'd use it again if the alternative was to have a too-long break between meetings.

Next time, let's super-quickly demo, at the end, where and how to edit the wiki, to see if that encourages more people to contribute feedback like this.

Aidan Budd 11:14, 22 March 2013 (CET)

best ways to collect our thoughts

I think if we follow the conventions then the talk pages are the best place to discus things (apart from face-to-face meetings). I don't like it when too many comments are inserted in the main page as I think it makes the real content hard to read. I like the idea of using these talk pages more than setting up a separate forum as it keeps the comments closer to what is being commented on. (One of Aidan's reasons for disliking using the talk page was that he 'wanted to make comments directly onto points already mentioned in the main page itself' - they would be even further removed in a separate forum.) Another reason I am reluctant to have a separate forum for these discussions is that it means yet another username and password for everyone. - Matthew 16:59, 28 January 2013 (CET)

Ah, thanks Matt, for helping out with that - all sounds very sensible! If I've time, I'll have a go at refactoring my comments from the previous page out here into specific headings, and reading the conventions on the link you give above more clearly - Aidan Budd 17:05, 28 January 2013 (CET)
I'm being hypocritical anyway - my original agenda was fairly comment-laden. I'll refactor too, after you've done the same. - Matthew 17:07, 28 January 2013 (CET)

I could install this forum extension for the wiki: [1]. Matthew 10:25, 30 January 2013 (CET)

Would be great to try this out - sounds great, Matt! Aidan Budd 20:43, 30 January 2013 (CET)
Just installed here. - Matthew 22:31, 4 February 2013 (CET)

HUB5 Planning meeting agenda (not sure which date this was!)

In general, how about we aim to restrict the agenda to:

  • choosing a topic, rough structure, ideal dates for HUB5
  • planning for a HUB4.5 meeting, perhaps between now and then, where we open up the discussion of:
    • topics for HUB6
    • what to do with flash talks in general (Maik has already proposed/volunteered himself for a flash talk for HUB5, I'm fine with just one per meeting)
    • ways to get the word out about the meeting to appeal to/bring in more people
    • ways of getting more pre- and post-meeting participation via the wiki, to get more of the community directly involved in shaping the programs i.e. that we do the decision-making at this meeting that is needed to get HUB5 happening ASAP, and work out what other issues we feel can wait a bit and thus which there is time to open up to a larger group for discussion i.e. HUB4.5

Also, general feedback on HUB4 - I think Marstall was a good venue for our after-HUB drinks - close, and with lots of available big tables; if they haven't got something on like a football game being shown, I suggest we go there again after HUB5; we could maybe even do HUB4.5 in there...?

- Aidan Budd 16:38, 28 January 2013 (CET)

I agree, Marstall worked well. - Matthew 17:44, 28 January 2013 (CET)


  • Morning meetings location: As an alternative location for the morning meetings, if you ever feel like changing, I would like to suggest Kaufhof in Bismarkplatz. I normally prefer proper cafes, but i checked it last week and for breakfast is a rather good option, it's very big, not so crowded, has different breakfast offers for 4 euro and a large table with armchairs where nobody usually sits. The view is great. On the down side it has no wlan. (suggested by Merche)

Invite others to planning meetings?

Once we have a date, shall we invite others by the mailing list? Several people put their hands up last night (HUB4) but I didn't take any names. - Matthew 17:38, 28 January 2013 (CET)

Don't mind (g.zipprich@dkfz.de), just need to check this page more often. - Gideon, 22:41, 28 January 2013 (CET)

potential topics for HUB5

I felt that there was a fairly even split between the two main ideas for HUB5 (and beyond), with a slight bias towards producing a game. There's a lot of cross over between the two ideas though, so would be cool if we could combine them somehow. Any ideas? Use the EBI DAS game as an example? Does anyone know any of the people involved? Perhaps they are in Heidelberg in March anyway? Some kind of networking / 'where do you stand on XXX?' exercise could be good for this - it was a bit chaotic last night (HUB4) but several people said they enjoyed it as a warm-up exercise and I liked how it got people talking. We won't have much time to produce something useful, so we would have to think about this carefully. We may need to come up with a general theme before HUB5. - Matthew 17:37, 28 January 2013 (CET)

I know both authors of the paper that was written to describe the game, Vicky will be here in May but can't join a HUB meeting, don't know if Rafael has plans to be here soon - Aidan Budd 16:22, 28 January 2013 (CET)
Really depens where we want to go. Just teach scripting (How to use BioPerl, BioPython...) or how to write clean code?
Good vs Bad Code
funny programming jargon
:D
If we decide to go for something larger than scripts, we should use github and for open source projects jira is for free. Also reviewing the code by other hub members would help for sure.
I somehow don't see the point of teaching scripting in a hub meeting, there are several other possibilties to learn it. We could try to solve "harder" bioinformatic problems (rosalind?) together using github and write the code inbetween hub meetings. In the next hub meeting someone could summarize what we achieved (number of problems solved?, everyone understands the code?...) and what difficulties we had. - Gideon, 23:11, 28 January 2013 (CET)
One comment here is that the ultimate and main aim (I think...?!) that we have at HUB is to get people in the community to know each other bit, by providing activities that are of reasonable interest to many of us, which involve us interacting directly with others - hence, from my point of view, while I'd like the activities we carry out to be useful, and to be perceived as useful, by the community (as I expect they'll only come to the meetings if they consider what's on offer is interesting), if the activities don't do a superb job at delivering what we hope they will in terms of that specific activity (e.g. improving our ability to communicate our science to non-specialists) **but** the activity still does a good job of bringing people together and helping them to get to know each other better, then I find that still a success.
More specifically, thinking about the fact that many of us don't perhaps need to be taught how to script - the focus, as I see it, isn't so much on "learning to script" but rather on doing some coding/scripting together with someone else, so that there is a context in which to talk to other HUB people about our code as we're producing it, and to give people a taste of pair programming, something many of us have not necessarily had experience of before. We could then, based on how our experience of doing something like this in a HUB meeting, think about extending the ideas to doing something more ambitious i.e. trying out some harder problems. Aidan Budd 10:52, 29 January 2013 (CET)
Sure, sounds reasonable. But what is programming? Is it: "how to use nice libraries (bioperl, biopython, pysam, picard,...) and solve the small problem as fast as possible?" or is it more "write code which every other programmer can read even if he is not that familiar with the language?" (in other words). Programming is also to look for tools already available and use them (let me google that for you). I agree with the main aim, just not sure if a 90min paired programming "hackathon" is the way to go. I just prefer solution 2: "Focusing on developing a (probably board, maybe later computer) game for helping wet lab biologists learning about some useful (for them) aspect of bioinformatics" - which does not exclude solution 1 for a later HUB#. - Gideon, 18:21, 29 January 2013 (CET)

Given that we've (I think...!?) now chosen to go with the "game" option, then here are a couple of comments on how we could organise it, and perhaps how it may be important to decide some things about the focus of the game **before** the actual event:

  • I think, before we do HUB5, we need to have already decided the target audience of the game; what constitutes a useful thing to teach someone **hugely** depends on this, so I think without this having been chosen, it'd be hard to do small-group discussions on other aspects of plans for a game e.g. choosing the form/format of the game, choosing what thing could be taught, ideas about how to teach it.
  • Thus, I suggest we try and use the wiki (or surveymonkey) to somehow collect opinions about the ideas of the community on this.

Aidan Budd 20:45, 30 January 2013 (CET)

I think this decision is something we can incorporate in to the evening itself, since we always get far more participation on the night than we do beforehand. And different game ideas will probably lend themselves naturally to different target audiences. The email etc to prime people in advance should include this in a list of 'things to consider'. Personally, I think we should aim for Masters / early PhD level biologists with not much bioinformatics background. If only because that's who I end up teaching the most, and so the game will be most useful to me if we go for them ;-) - Matthew 23:24, 30 January 2013 (CET)

dates for HUB5

Depends on room availability and speaker availability. - Matthew 17:37, 28 January 2013 (CET)

Intro to bioinformatics at a HD institute

Anyone we could ask now for March? - Matthew 17:37, 28 January 2013 (CET)

is now in Minutes, copy paste or move minutes into Talk:HUB5? - Gideon, 17:48, 31 January 2013 (CET)

Roland Eils or Benedikt Brors from DKFZ - if we have a date set for HUB5 I can ask them on the eilslab retreat which will last from 03.02.13 to 08.02.13 or aim for a later HUBX for another Introduction? - Gideon, 19:08, 31 January 2013 (CET)

I asked Roland and he would do a presentation. Maybe we should settle the topic better, many of us already know what bioinformatics is at the eilslabs, or not? I did not asked for a specific date so once we have a date for HUB5 or HUB6 I would ask him again. - gideon
Thanks Gideon. Thursday 21st is the provisional date for HUB5. Jon has provisionally reserved a room and is waiting for it to be confirmed. Then I'll send an email to the organisers and afterwards to everyone, though I'm away next week. (I moved this discussion to the talk page as I think the minutes should only list what was discussed at the meeting, not anything that came after.) - Matthew 11:25, 8 February 2013 (CET)
Regarding the topic - I think an intro to bioinformatics at DKFZ / in the Eils lab will be good for everyone who is not at DKFZ; 'Introductions to the large bioinformatics-employing organisations around Heidelberg' was the most desired session in the survey after HUB1 - see Q19 on the survey page. - Matthew 11:30, 8 February 2013 (CET)

Flash talk(s)

Maik has already offered. Maybe one is enough? - Matthew 17:46, 28 January 2013 (CET)

Adam suggests flash talks themed around HUB3 biggest challenges in bioinformatics, eg. starting with the 'what is a species?' topic:

  1. 5 mins: Speaker 1: 'Simple, we can tell by the DNA'
  2. 5 mins: Speaker 2: 'No! It's not that simple'
  3. 5 mins: Audience debate.

Topics to discuss at HUB5

(Moved from the main page to tidy it up a bit - Matthew 15:52, 4 March 2013 (CET))

  • Preparing a bioinformatics-focused game
    • First steps in board game design
    • Purpose?
      • Education
      • Advertisement (of HUB, or of a particular field of bioinformatics, or...)
      • Entertainment
    • Audience?
      • Children
      • Teenagers
      • Adults
      • Scientists who are not bioinformaticans
      • Bioinformaticans
    • Possible Subjects:
      • DNA -> RNA -> Protein
      • Sequence -> Structure -> Function
      • Networks

Game ideas

Food for thought:

Science Game Center

Paper about DAS game we referred to in our discussion previously Science Game Center - copy pasted from Aidans email, ty! - Gideon, 22:43, 31 January 2013 (CET)

I worry slightly, that we're trying to do too much in a first meeting on this topic (i.e. the HUB5 meeting itself, trying to develop a game), by aiming to go all the way from 0 to choosing already a concept for a game and even drafting it on that evening.
I think a key part of the success of a game we develop, would be that it teaches something that many of us (and potential journal editors) consider important.
I also think that, to be of general interest for HUB participants, ideally the thing we teach is something fairly generic/central to bioinformatics, something we all (or many of us) need to understand
Thus, I think it'd be good to devote quite some time (world cafe again...?) to collecting possible ideas about what the learning objective(s) we'd like to focus the game on. This could be done in parallel to choosing a format of the game, as many formats can be used with different kinds of questions.
Also - by focusing on the learning objectives we kinda "automatically" are selecting our audience for the game i.e. the people for whom this particular laerning objective would be directly useful/appropriate
I think that learning objectives that are more specific than just "learn how BLAST works" would be useful
Do any of us know people who develop games professionally? I think it'd be great if we could get some feedback from such people about key things we could/should be thinking about while putting a game together
I'd be aiming to use the HUB5 as just a first step towards developing the game, using the time to exploit having lots of people and ideas around, but aiming to finish the project off later maybe in a smaller group, maybe involving also a later HUB meeting, etc.
Aidan Budd 18:20, 6 March 2013 (CET)

Some of the game ideas below are 'just' biology, not bioinformatics. I don't think this matters though. Do you? - Matthew 21:46, 11 March 2013 (CET)

Some of the game ideas below are copies of well known games but with a biology/bioinformatics twist., They maybe be easier way to get something that works, since the mechanics are already established, but might not be that exciting. Dunno. - Matthew 21:46, 11 March 2013 (CET)

Ice-breaking games

I don't know what you have in mind as "warm-up" games for the meetings. In CS ice-breaking games are often used in big events so people meet each other. Depending on the group size they take 10 to 30 minutes. One of the classics is to give each participant a "card" with country flags. The task is basically to find someone from each one. You could come up with something similar, setting up a task in which people have to find at least one person from each participating institute; or working in X (biostatistics, NGS, ..) ; or else a card with 3 or 4 questions where you challenge people to go around the room and find someone else that can answer them. The first one to complete it wins, and maybe earns a beer ;) I'm not sure this would make sense with less than 25 participants though

added by Merche

Hi Merche, sounds good. Previously we've done things like 'where do you stand on X?', and got people to place themselves on an (imaginary) map of Europe to represent where they're from, or along an imaginary line to show where they think they fit on a bioinformatics scale of pure biologist to pure computational scientist. Another example we've used is to get people to come up with elevator pitches to quickly describe their work to someone else. Any and all ideas are welcome, the more variety the better I think, to keep things fresh. They need to be fairly quick though, as we usually only have two hours for the whole meeting. It'd be great if you would add your ideas to the suggestions for future HUBs page. Thanks, Matthew 14:00, 19 March 2013 (CET)

Overall structure of the Game

I agree with Adam above [Do you mean Aidan? - Matthew 13:12, 19 March 2013 (CET)] that you might be aiming too much stuff at once and I think it would be best to have a preliminary idea of the game in mind before the next HUB meeting. Why? Simply because if you ask 25 people ideas about a game, you will probably get ideas for 25 different games, not 25 ideas for a game, which is what you actually want. Besides it would take many HUB meetings to come out with an actual game. My proposal is this:

I would go for a Party & Co. kind of game, mostly because:

  - It is sort of a compilation of many other games: Tabu, Pictionary, Trivial, .. The format suits quite well. We can import some of the original game categories or add our own
  - It is meant to be played on teams

We can pre-select four or five categories and create one discussion group for each, asking participants to provide questions or ideas according to their experience and background or their lack of it. For example, if everyone provides:

  -one "advanced" question on their background
  -one on a topic they have basic knowledge about 
  -a last on something they are interested in, but not necessarily know the answer to

I think this would ensure the interdisciplinarity of the game as well as different levels of complexity. In order not to get too much off-topic, it would be a good idea for us to provide 2 or 3 sample questions. The modularity of the game would also make it very easy to integrate all game-ideas bellow and add/change a category if someone has a good idea or something does not work. On top of the category, we could add a last challenge for the first group (or groups) that reach the end of the game. Here we could integrate Aidan's idea of a debate-like challenge. Having all this in mind, we must be aware that the game should be sensible, we don't wanna come up with a monopoly-like game that can last for a week

Merche 11:15, 19 March 2013 (CET)

Rather than having ca. 25 different games, the current idea (see the minutes of the last planning meeting) is to have a voting round at the beginning of HUB5 where the most popular ideas will hopefully rise to the top, with the top four or five then being worked on in groups. The groups will then present their ideas to everyone at the end. This is hopefully in the spirit of the intended participant-driven nature of HUB in general. The aim then of HUB5, as I see it, is to get some ideas of games that will work, some idea of how, and a measure of people's enthusiasm to take them further. - Matthew 13:12, 19 March 2013 (CET)

Protein network

'how about "jumping along a network" of proteins, could be related to diseases or not. you only "see" the next immediate neighbors and must avoid hitting a dead end (game over) the more proteins you manage the more points. This way a player "might" learn about proteins, names interactions etc.' Suggested by Holger. - Matthew 10:18, 30 January 2013 (CET)

The edges could be different types of interactions? Direct and indirect, binary and other (TAP complexes, etc). With proteins with domains and linear motifs, with interactions mediated by those. - Matthew 10:18, 30 January 2013 (CET)

Protein modularity and interactions

To teach principles of protein modularity (structure domains, disorder, linear motifs, signal peptides, transmembrane domains)... and, additionally, how they relate to protein-protein interactions. Cards / pieces could represent different domains, with each player being dealt a selection at the beginning, from which they construct their protein. Other cards then describe which domains can interact, and the players compete to define the biggest network. Perhaps homology could also be covered somehow, through the proteins of different players sharing a domain or domain type. - Matthew 21:46, 11 March 2013 (CET)

Signal transduction / Infection of a Cell

Compete to get your signal from outside the cell to the nucleus in the shortest time, or to override opponent's cell's defenses first [2] . This came up in our (Adam, Aidan, Holger, me) discussion with Philipp Gebhardt - I may be missing something... - Matthew 21:46, 11 March 2013 (CET) Matthew

'Taboo'

Given a concept on a card, explain it to the other players without using the word itself [3]. Suggested by Gideon. - Matthew 10:18, 30 January 2013 (CET)

you can also expand it to pictionary (e.g. paint NGS or WesternBlot.. or even use pantomime (probably more feasible for wet lab experiments) this would more go into the direction of the game Activity (German Wiki, boardgamegeek, blog post about activiy, good introduction - Gideon, 20:08, 31 January 2013 (CET)
relies on the player knowing what the word / phrase means, which won't necessarily be true, especially if the general idea is to teach some aspect of bioinformatics / biology. - Matthew 09:36, 4 February 2013 (CET)
could be combined with Up Goer Five. (I think it was Jon and/or Holger who suggested this originally?) - Matthew 21:46, 11 March 2013 (CET)

HUBris

Similar to taboo, except one should describe the given word or phrase by making grand, exaggerated claims about it. OK, I'm just looking for ideas with 'HUB' in them somewhere. - Matthew 12:30, 3 February 2013 (CET)

relies on the player knowing what the word / phrase means, which won't necessarily be true, especially if the general idea is to teach some aspect of bioinformatics / biology. - Matthew 09:36, 4 February 2013 (CET)
Sure, but for example you could make 2 groups one more bioinformatics one more biologist, each group has one stable of cards which they know and it should be possible for them to explain it with words and/or painting to someone with a specific amount of basic knowledge. In the end it comes down to whom you want to teach something. If the game should go more into the to teach something to someone where you have no or nearly zero knowledge about then activity or taboo is not optimal. - Gideon 17:12 CET typing this using my smartphone so feel free to improve it.

Top Trumps

A set of cards with various facts about proteins (for example), with players competing by selecting a fact on their current card that they think will beat the same on their opponents cards [4]. - Matthew 09:36, 4 February 2013 (CET)

Memory

Cards with protein structures, for example, that should be matched. Like common memory games for children. Suggested by Holger. - Matthew 10:18, 30 January 2013 (CET)

eg. protein stuctures etc. maybe from different angles to make it more complicated? (Suggested by Holger)

Programming challenge

Provide a (very) short piece of code in a card so people can guess what it does (Java,C,C++,R,perl,python, mySQL, ..). If time allowed is reasonable, ask the group to come up with a small code to complete a (simple) task. Or experienced R users could come up with something like "Suggest three R packages most useful for X ", .. I leave it open for suggestions :) (Suggested by Merche)

what are the most popular HUBs? are different types of people attracted to different topics?

Can we attract techies, biologists, and interested lay people to the same HUB somehow, or are we better varying the degree of each between HUBs? More people seem to attend the more science-based HUBs I think (though this may be just coincidence - I haven't calculated statistical significance...) - Matthew 17:47, 28 January 2013 (CET)

I also have the feeling that more people are interested in attending the more technical/science-based HUBs. Meh. Not sure what to do about that, as it seems almost counter-intuitive to me that more people will tend to attend the more specific (and hence I would have assumed of less broad interest) topic-focused HUBs...? But maybe we need to try again, next, something more techy... Aidan Budd 19:18, 20 March 2013 (CET)

HUB5 task list

I've removed the drinking water, beer and pub booking tasks since we'll already be in a pub (restuarant) and most probably can't bring our own things. - Matthew 15:54, 4 March 2013 (CET)