We are in the process of creating bioinformatics mobile applications. Rather than common app we want to give app for scholars and scientist for them to access the data wherever they and whenever they want.
Please give your suggestions and recommendations to pick the area or functionalities need to be implemented.
This and the related thread is getting a bit out of hand in that they are getting too personal. I would suggest that we all take a break from it for a few days.
I wish the authors success with their app.
Regardless of what value someone might associate with their product at this stage it really seems that the authors have good intentions and are just inexperienced in this field. Perhaps, rather than posting a followup the authors might just want to go back to the drawing board and reflect on why people's perceptions of this product ended up so different than that of the creators. And in the future try to avoid the path that lead that way. There are many lessons there.
In Ensembl we're currently looking into a mobile website (not an app) and are asking ourselves similar questions. We don't envisage anybody doing proper analysis on a phone as 5heikki has already stated, but we imagine something more like a quick gene look-up when chatting over coffee or in a conference. If anybody does or has tried to look up this kind of data on a phone, we'd be interested to know about the kinds of things you were looking up.
ADD REPLY
• link
updated 2.6 years ago by
Ram
44k
•
written 9.8 years ago by
Emily
23k
0
Entering edit mode
My lab would benefit from a quick look for a gene, its variants with effects predicted for each variant - something like HGMD, but for mobile devices. Any serious discussion after that is gonna happen on my Mac.
Hi Emily, this is indeed how I use most often my smart-phone and tablet for biology: getting more info about a molecule/structure/disease etc. during seminars or even group meeting.
I think this is a counter example to implementing a software.
It is very unlikely that you could implement a useful tool if you don't know what people need to begin with. You should have to have a basic utility and mode of operation in mind and the feedback that you seek should be on adapting and tuning that.
We have N number of plans from basic view to mutation pipelines. But what wanted to understand who needs what. It is always good to understand what people need.
There are quite a few online pipelines, but what is the benefit of mobile apps here? I'm pretty sure any bioinformatics person that can operate a mobile app already is using VPN to connect to their high power work machines and/or clusters/HPCs.
Ram, I can understand the pain point you are having from your previous experiences. As most of them said here, UIX matters. Can you please let me know what are the pipelines are available at present?
Here's my problem: you've shown and now admitted that your knowledge of bioinformatics is lacking. At the same time you're peppering us with absurd promises and asking us to invest our time in a product for whose existence we don't yet have real evidence. Your very attractive template-based website that slickly obscures any specifics of the underlying product only furthers this perception. I especially like the pineapple non-sequitur at the end.
I'm thinking you're trying to get seed money from some person or organization by roping a few of us into signing up for your platform and using those signups as evidence of involvement and interest from the greater bioinformatics community. Maybe I'm wrong, but so far this whole thing seems to be a waste of our time and possibly fraudulent. Thus I feel compelled to call you on it.
Underwhelming performance, a tiny display, plain awful input.. yeah bioinformatics apps on mobile devices sounds like a great idea. A very creative person could maybe come up with useful companion apps to real programs.. other than that. No.
The only time I've used a mobile device (I used a tablet, not a phone) to do anything related to actual bioinformatics work was when I SSH'd into my HPC and ran a qstat. Most of my work is done in the terminal and almost all of it involves Control keys (Cmd/Ctrl/Alt/Opt/Esc/Tab) - IMO a mobile interface will only serve to slow me down and irritate me.
Mobile interfaces are amazing when an application is UX intensive, but for data intensive/pipeline oriented tasks, I highly doubt any app would help.
TL;DR: Plain text manipulation is a nightmare on mobile devices. So, mobile apps are not really helpful.
I'm here to look for the feedbacks and need of industry. Be open minded, can you tell me single tool, which has rich in terms of UI and features for Bioinformatics and affordable by everyone.
For us its all data and process with subject matters experts. if you don't know let me tell you, we build a pipeline for a customer. Which was seen as a challenge some of big players. If you don't know or not open to share your thoughts it's good to stay away.
Are you open for the challenges? We are and we will deliver if we are committing something.
You realize there's a reason why most bioinformatics tools don't have a flashy UI, right? I cannot help it if you equate flashy GUI to "good UX". They are not the same. UI should work the way users want it to.
In our case, automation and programmability takes precedence over graphical toggles. We're not interested in typing bwa commands or picking BLAST options off a UI 400 times a week or reading through pages and pages of GATK's best practices.
We do it once, write a bunch of make files that do this stuff for us and then write a bunch of configuration files that inform the make files. To be able to do that, we have to stick to institution specific idiosyncrasies - such as the HPC system, the batch schedulers, the environments, user auth etc.
While a website for the topmost layer would be useful to a high throughput lab that handles an entire organization's internal and external requests, I do not see individual users benefitting from a low-customization environment. And while I can automate/copy-paste+modify commands, I cannot do that with app interactions as easily.
That aside, now that you have committed to the de novo ultra short read assembly with an API for low coverage data (oh the horror - Dan D , you're truly magnificently evil), I'd love to see how you're going to deliver that.
Guys, NEVER agree to something you do not understand.
Yeap Dan,
We can enable this feature. The same can be hosted both in private or public mode. Please write to support@bioelm.com with more information on you need. We are happy to enable this feature for you.
I'm really stoked that you can do this! I always thought it was impossible. In fact I was sure of it. Now that you've shown just how amazing your software is I'm sure more people here will have their own big challenges.
Dan,
Why don't you share few more information on this and we can work together? Please share you thoughts. aeinsights is one of the developer from our team. Thanks.
This is really the problem with your approach, you don't actually know bioinformatics.
Dan has stated a nonsensical problem to see if you can tell nonsense from actual science.
When you said yes we'll get right on it and that you can enable the feature right away you kind of tell us that you don't really understand this field.
But I do actually have a serious proposal that would make me drop anything and seriously
investigate your services and become a fan.
Pick a published bioinformatics analysis from a high impact paper. Pick a glam journal: Nature, Science that gets people's attention.
Reimplement their pipeline in bioelm. Starting from raw data have it produce and visualize some (at least one) of the main findings of this paper.
Demonstrate how users now can do this analysis from their phone.
Have a push notification that says: "Nature Paper reanalysis completed in 5 minutes." now that would get my attention ... :-)
If you do this I'll be your fan. I am sure many people here would love to give you suggestions which paper to implement.
Guys,
I just want to understand why this is not feasible. Can someone please answer? I'm not a subject matter expert of Bioinformatics. Why we can't access FFPE data and expose them over API?
You can access FFPE data and expose it over an API - that's called file hosting. Try hosting a de novo assembler. Try designing one that works with short reads, let alone ultra short reads. Then try dealing with low coverage.
Exposing files over API is so simple that it takes maybe 10 minutes to do that.
Yeap I'm not expert in bioinformatics but I can follow when subject matter explain me what it is all about. That's the reason I'm more eager to get beta customers, who can start using our platform.
Sure will make you as fan of our platform... so soon.. :)
Dan, Can you share the data if you have, we will host them for free with graphical representation of data and make it available for both web and mobile. Thanks.
Guys, honestly, if you're looking to build credibility, you're going about it the wrong way. This latest commitment misstep has pushed your claims from "unbelievable" to "highly ridiculous". That's why people are having some good natured fun at that expense.
The FFPE data isn't even the main point.. The ultra short and low coverage just doesn't play nicely with denovo assembly. If you can do it, then I guess we have been wasting our time in bioinformatics for the past decades....
I use pubmed on mobile sometimes. Other than that may be querying population frequencies for a mutation locus might be useful but otherwise I can't think of much else I'd want to check on my mobile.
Guys,
It's great to see lots and lots of interest from the people over for mobile app. We build a platform that will enable quick view of your data with alters and notifications. Please visit bioelm.com for more info.
It appears to be blocked within my university network (chrome & firefox on desktop, chrome on mobile via eduroam) but works on mobile chrome when disconnected from eduroam. Don't know if that is of importance.
This and the related thread is getting a bit out of hand in that they are getting too personal. I would suggest that we all take a break from it for a few days.
I wish the authors success with their app.
Regardless of what value someone might associate with their product at this stage it really seems that the authors have good intentions and are just inexperienced in this field. Perhaps, rather than posting a followup the authors might just want to go back to the drawing board and reflect on why people's perceptions of this product ended up so different than that of the creators. And in the future try to avoid the path that lead that way. There are many lessons there.