colin-combe / crosslink-viewer Goto Github PK

Hey guys,

I can't get the upload functionality working on my own server. When I try to upload an annotated csv file I get: "No CSV file uploaded." And when I try to upload a csv file I get: "Could not connect:" without any further information.
Apart from that, I connected the php files like described in the how-to.txt.

This is how my database table "upload" looks like:

Any idea whats going on?

Thanks!

needed, for e.g., when selecting multiple links to show in 3d view

related issue #11

refactor readCSV so initialisation function calls are outside it.

Aim is to let readCSV be called repeatedly, allowing to add data from multiple seperate CSV files

always use same colours for annotations

Cache sequences and features in local storage (data from uniprot and superfamDAS). Its better than downloading them every time.

Don't use the uniprot DAS server,its getting switched off in 2015. We will use the text uniprot web service, e.g. http://www.uniprot.org/uniprot/P12345.txt

There is a JSON uniprot service (http://wwwdev.ebi.ac.uk/uniprot/services/feature/Q94PW3.json) but it is incomplete and still under development.

see MICommunity/ComplexViewer#75

@juanzou - need to greyout other peaks when 'highlighting' one, other wise too cluttered

Be able to hide individual cross-links.

If things have been hidden then it should be made clear, so user knows they're not looking at whole dataset.

feature request from @abelsom

needed by https://github.com/colin-combe/CLMS-UI/issues/5

implement touch screen events

homo-multimeric links are not shown on the opposite of the protein and not in right color.
Also the class gets not correctly set/reset (e.g. bendiness) on filtering matches.

bend / break protein bars and move sub-parts around in 2D

example can be seen in http://www.sciencedirect.com/science/article/pii/S0092867413010106

feature request from @abelsom
one of the more tricky things to do, largely requires rewrite of viewer but this is becoming more and more necessary now

When showing multiple searches, provide option for whether to compare data sets (using different colour for each search) or just aggregate them (don't distinguish searches, leave colour for e.g. distance).

Feature request from @abelsom

read tab separated values, some users have their data in this format. Ideally, make an intelligent guess as to whether data is CSV or TSV

Right click allows positioning of the crosslinks from top to bottom of the protein bar.
But this change is lost when lay-out is exported and then opened in adobe illustrator, making figure cluttered.

The sequence data comes from the uniprot .fasta service, the features/annotations from the .txt service. This introduces possibility that the features shown are not correct for the sequence. See #2

Need to at least highlight this possibility, or not show features for iso-forms. A user suggests:

• to have an additional "mixed annotation" option (if the custom domain annotation is not present then the annotation is taken from the databank)
• to display some kind of warning that a particular protein is not a default isoform (therefore the domain annotation can be wrong)

Use SignalJS for event handling

the events branch has comparison of datasets and non unique self-links are shown on other side of bar.
However, they do not always/ever flip back if they become unique due to filter changing.

It seems there is currently no cross-browser solution for client-side file download.

For the Safari issue see:
eligrey/FileSaver.js#12 (comment)

Demo 2 shows domain data. Demo 1 does not.

As with interaction-viewer (colin-combe/ComplexViewer#2) there seems to be a problem with the DAS server not returning sequences for proteins isoforms.

We may need to change to using other uniprot web services.

theres currently a problem with this, caused by the somewhat complicated initialisation process for Proteins (which is in turn caused by the fact that sequence may not be known when loaded)

Seems its broken in IE due to use of .contains()

Same issue as MICommunity/ComplexViewer#78