search

PitchInteractiveInc / Phinch

Phinch is an open-source framework for visualizing biological data, funded by a grant from the Alfred P. Sloan foundation. This project represents an interdisciplinary collaboration between Pitch Interactive, a data visualization studio in Oakland, CA, and biological researchers at UC Riverside.
http://phinch.org/
BSD 2-Clause "Simplified" License
149 stars 30 forks source link

Old files will load but newly created json BIOMs will not load #74

Open Skolz3869 opened 7 years ago

Skolz3869 commented 7 years ago

Hi, Phinch is great. Fantastic way of getting people to take ownership of their data.

However over the last week I have had issues getting new files to open. I can get previously uploaded json BIOMs to load, however any new BIOM's (also json formatted) sit on the filter data page with the circle rotating and nothing happening.

Is there an issue with the server? Thanks, Emily

hollybik commented 7 years ago

Without seeing your files I can't diagnose the issue - however, I would note that some users have reported problems uploading BIOM files prepared in QIIME v1.9 and higher (we're not sure what is going on here, and we're hoping to find the issue when we resume development this summer). Make sure you're using a BIOM v1.0 file (check the header), and I would also check that your taxonomy and sample metadata are both added correctly to the file. Any of these things can prevent files from loading correctly.

Skolz3869 commented 7 years ago

Hi Holly,

I think that it may be the QIIME 1.9+ issue. I have almost completed a review of web services which process / analyse / visualise 16s data. I would love to include phinch, but can't get it to work on my test data set. This is despite it working on previous sets.

Glad to hear that you are going to do some more development. I love this program, it is really useful for getting non microbiologists involved in data.

One thing that is missing in terms of visualisation among all the web services is the ability to have a diagram like below. The data is ordered by metadata category (in this case CST) or date and the top 10 OTU's (or whatever taxa) are down the side. They are colored by relative abundance. Many programs give you the ability to create much more complex diagrams with clustering, or where complex normalizations have been made etc, but are missing the simple and effective representation like below (made in Primer 7).

In this case it is clear that CST does not appear to have a link with biological populations. I have found that this is the best way to represent time-series data too.

[image: Inline images 1]

Thanks once again for your awesome program.

Regards, Emily

On 1 June 2017 at 23:23, hollybik notifications@github.com wrote:

Without seeing your files I can't diagnose the issue - however, I would note that some users have reported problems uploading BIOM files prepared in QIIME v1.9 and higher (we're not sure what is going on here, and we're hoping to find the issue when we resume development this summer). Make sure you're using a BIOM v1.0 file (check the header), and I would also check that your taxonomy and sample metadata are both added correctly to the file. Any of these things can prevent files from loading correctly.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/PitchInteractiveInc/Phinch/issues/74#issuecomment-305491755, or mute the thread https://github.com/notifications/unsubscribe-auth/AWACH4cHS8mXmJpAsUud6q-Z8Xiik_UCks5r_rthgaJpZM4No6pr .

-- Emily Scholes

PhD Student SASE & Gippsland Water Factory Mobile: 0434210610 Monash Office: (03) 99026202 GWF Lab: (03) 51775960