What Do You Think Microsoft Biology Foundation?

Question

[Sirus]

Hello every body, I have seem that Microsoft has published the Microsoft Biology Foundation library which seems very interesting, even if it is limited for the moment, as far as I have seen it implements only sequencing algorithms and different file format parsers. but in addition it gives some interfaces to enable us create our own algorithms. I wondering did any one use this library and what are the fields in which we can use this library？can we use it for example for protein-protein interaction networks? How about the memory consumption? as I have seem it seems that it uses a compression technique to store data in the RAM.

Can any one share his experience, because I am a .NET developer and working in bioinformatics I am very interested about this library.

Thank you in advance.

Answer 1

[Rick Benge]

I am the community program manager for the Microsoft Biology Foundation. I am not a Microsoft employee but work under a contract for them. If you have any specific questions I am happy to answer them directly or via our discussion forum's hosted on Codeplex. This is an open source project that is being utilized within Microsoft, however it is open to the community for contributions. We just released in late April the version 2 beta1 release. As to whether it works on Mono - we don't specifically test on that platform, but there is nothing that should prevent us from working at this point. If you have information to the contrary I would like to see the details. You can post an item under the issue tracker tab on Codeplex. MBF is developed upon .Net so you will see the system requirements are partly a result of that dependency.

Here is the link to Codeplex for your own reference http://mbf.codeplex.com/discussions

Rick Benge Community Program Manager – Microsoft Biology Foundation | Microsoft Research Connections | Office [phone]

[lh3]

It is extremely important to make sure the library run on mono, which is the only way to make it available to researchers in the academic world. Have you published MBF, BTW?

[ndunofficial]

Just in case anyone is seeing this years later, the code base now runs on linux and/or windows. The main developers are evenly divided among the OSes that they use.

[Michael Schubert]

Couldn't agree more with lh3. Given the current state for Windows vs Linux in bioinformatics, it would be a waste not ensuring mono compatibility.

[Sirus]

Nice to have this info me too in fact I didn't try it on Mono yet, I'll try it soon

Answer 2

[2184687-1231-83-]

I think it's a good initiative that hasn't picked up much momentum. As far as I have seen in publications, there are a couple of really good .NET bioinformaticians around, but the hordes of Java bioinformaticians haven't switched to .NET en masse, so a lot is happening in the Java-world instead...

[Sirus]

Yeah, right, I think it can be used at schools for teaching or for quick solutions

[Egon Willighagen]

Indeed. Most of my critical code is in Java, and until I can easily use those Java libs in .net, I cannot switch.

Answer 3

[Andra Waagmeester]

I wasn't aware of this initiative. But when I checked, I saw the following requirements:

System Requirements

Windows XP with Service Pack (SP) 3 or later
Microsoft .NET Framework version 4.0

I was under the impression that OS-specific development was something more or less of the past (Except of course the Mac specific cocoa developments, which I ignore for that reason). If this MS library only suits the need of MS OS's, it is a no go for me.

There are to many Linux and Mac machines out there to just ignore them. If the MS biology foundation library is about ignoring the other platforms, I think that it is not worth considering.

[Neilfws]

I'm unable to answer this question objectively, because I have a Pavlovian response to the word "Microsoft" (and it is not salivation). You make a good point - Microsoft are not interested in cross-platform solutions. Their number one goal is to make you use Windows, not to develop good research tools. And that's why I simply stop reading when I see the word.

[Michael Schubert]

Neil, I do not agree with your general condemnation of Microsoft Research products as they are way past just trying to "make you use Windows"; if you dig a little deeper you see that it's Apache licensed and quite Mono friendly: http://www.infoq.com/news/2010/08/Mono

[Eric Normandeau]

Just having 'Microsoft' and 'Foundation' in the same phrase makes it feel like a scam :P

[Andra Waagmeester]

It is always nice to have -1 without explanation.

@michael The system requirements mentioned above are from http://research.microsoft.com/en-us/downloads/304eb276-8ed2-4cf9-ae1d-e48c38ba94b5/ It is MS themselves that states that you need windows

[Michael Schubert]

Windows XP is listed because .NET 4.0 is not designed to work with 2000 and older and Microsoft does not officially support Mono so that's no surprise. In general (and with a grain of salt), .NET works on Mac and Linux as well.

[Sirus]

For linux platforms one can use Mono, it is compatible with the .NET framework 4.0, I think it will not be a big problem

[Neilfws]

As I say, this is an issue on which I cannot be objective or dispassionate. The very word "Microsoft" just sets me off. Sorry :-)

[Egon Willighagen]

I'm at a MS workshop right now, running Linux... but there is some cool Open Source stuff, and F# just have linux binary packages... they're improving... (but I haven't received a phone yet :)

Answer 4

[Vince]

Hi sirus,

This answer will provide mainly my thoughts on the MBF and may not answer your question directly, although I hope you find it informative.

I was a summer intern who worked on using the MBF (v1.0). I developed a software program that you can find here. I had no C# or .NET experience prior to my internship and could have been considered a open-source enthusiast (not anti-MS, but a Python enthusiast in favor of cross platform compatibility and open-source initiatives).

At the time of my internship the functionality was limited, but is understandable considering the age of the project (about 1 year). MBF v2.0 beta is a major step in the right direction and seems to supports large data sets from next-gen sequencing. I have tried to compile the MBF (v1.0) using Mono and was somewhat successful in Windows (with help) link. I plan to try a Linux compile soon.

The success of the MBF will depend largely on how the open-source/bioinformatics community embraces it (the corporate environment is also important, but I will not consider it for brevity). Considering that many researchers with rudimentary bioinformatics experience use the Windows platform, developing bioinformatics software with the MBF makes sense as it caters to their usage habits. However, this is not to say that it is locked into this platform (see above). I envision that a Linux and Windows compatible MBF is possible with the communities' help, and would benefit both software developers and researchers in that the former can develop tools used both for the server (Linux) and desktop (Windows) environment and the latter will benefit from easy-to-use programs. Also, considering that Silverlight is somewhat cross-platform (Moonlight for Linux), GUI applications can also be cross-platform.

Personally, I enjoyed using C# and .NET and found that this alone was enough to get me interested in using (and continue to use) the MBF. I believe it has potential and should be assessed objectively and not based on prejudice, preconceptions or Pavlovian responses.

[Sirus]

I really appreciate your answer for many reasons. first it was from a non .NET guy and second is I like your analyse, I think for the long term BMF will be a competing library in the bioinformatics field and may be it will be as competitive as Bioconducter library who knows. Yeah, for the moment the cross-platform option is not that obvious but in my opinion it is a matter of time and as you stated the library is still mature.

[Vince]

Thanks. There are some training materials for the MBF at: http://research.microsoft.com/en-US/projects/bio/training.aspx. I have not used them because they were made available after my internship; however, I have attended presentations in which they were used and I was impressed by the example usage cases; they touched everything from simple manipulation of biological sequences all the way to building a simple GUI application in WPF.

Answer 5

[Rick Benge]

Small update here. MBF has been renamed to .NET Bio 1.0 and the codeplex site is now http://bio.codeplex.com/. I did list some details re: Mono under the FAQ. Would love to get more feedback on Mono usage. Also we will be hosting a training course in Redmond on Dec. 5th and 6th if you want to hear/see about this in person. You can register for the course at http://dotnetbio.eventbrite.com/.

Rick Benge .NET Bio 1.0 community program manager

[Egon Willighagen]

It was demoed yesterday in Stockholm, but unfortunately, the first half of that presentation was really having a strong flavor 'we do not suck'. I'm convinced of that, but it was not good advertisement. License is good, but .net is just not a popular language. I rather see a "F# Bio". I hope the teams will team up, and then they'll have something convincing.

Answer 6

[Nibua]

I think I could be a good solution for enterprises in bioinformatics. The .Net allows to easily set up interfaces. Then if it's also easy to provide common bioinformatic tools they would make a point (to me). But for researchers I'm pretty sure that other technologies that runs on Linux, Mac and even Windows systems will be more widely used.