New analysis for MUX table

[East-YutangChen]

Hi all,

I have developed a new analysis script which can be able to process the mux table. The KSEA_MUX.py and KSEA_MUX.ipyn is now on my fork and there is also an explanation pdf for the codes. Please take a look.

In order to handle mux data, my idea is to transform the original mux table to a n x 7 column table (see below) and add an extra Inhibitor column to index the phosphoproteomics data from an explicit inhibitor. So this means that we can subset the transformed table to get each inhibitor's table and then we can apply our analysis to each inhibitor. The problem here is now we need to change the web template so that we can implement the analysis for mux data in our web app. One of my idea is when we click submit button on the homepage it will direct us to a new page where we can show the name of inhibitors and by clicking the inhibitor we can go to the analysis results page for the inhibitor we have just clicked. Another idea is instead of having only one submit button, we can have an extra inhibitor or analysis button which can be dragged down to show the inhibitors and by choosing one inhibitor, the web goes to the results. What you think. Sorry for causing trouble for you guys.

[AnnaDearman]

I am so amazed by your functions!!

[JuanLM1978]

Guys you all are very smart!

Please dont forget that you also need to write an individual report ok?

[AnnaDearman]

We are aiming to stop working on the website by the end of Tuesday so should have time

[AnnaDearman]

Hi @East-YutangChen ,

I've been incorporating your script into the application.py script. I'm so impressed 😅 and I'm learning as I read it!

There is a slight problem that I spotted when trying to test your Jupyter notebook before incorporating it into the script. The line "df1 = df_transform[df_transform.Inhibitor == inhibitor]" throws up an error ("Lengths must match to compare"). I integrated it into the script anyway, in case it was a problem with my version of Python or something, and once I'd fixed a few of my bugs from the integration process, it started throwing this same error again, so I can't proceed.

What is the line of code trying to do?

Anna

[AnnaDearman]

I just checked your WhatsApp screen shot again and I can see that you have to index one of the inhibitors in inhibitor e.g. inhibitor[0].

I will ponder this! I will almost definitely need @celiaccb's help! 😄

[East-YutangChen]

Hi Anna，

Sorry for the delay. You are right, we have to enter the name of an inhibitor so that we can select the table of the inhibitor from the df_transform and then do the following analysis.

Kind regards,

Yutang

在 2020年2月9日，16:41，AnnaDearman notifications@github.com 写道：

 I just checked your WhatsApp screen shot again and I can see that you have to index one of the inhibitors in inhibitor e.g. inhibitor[0].

I will ponder this! I will almost definitely need @celiaccb's help! 😄

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[AnnaDearman]

No worries! I'm making progress! The app is producing a plot and a table for each inhibitor, I just haven't put downloadable links to them on the screen yet.

I've also included a couple of lines of code (if statement) to deal with any files that may be like mux but without the CV columns. I will test this soon.

Will you be working on making the displayed RKA table(s) show only the top and bottom 10 kinases by log fold change, and on making the bar plot be sorted by log fold change?

[East-YutangChen]

Actually, you don’t need to add extra codes to deal with files without CV columns because I have included these codes in my functions. If you have time you can try to sort the RKA table to show the top 10 subgroups and also the bar plot thing. I don’t mind.

2020年2月9日 下午5:08，AnnaDearman notifications@github.com 写道：

No worries! I'm making progress! The app is producing a plot and a table for each inhibitor, I just haven't put downloadable links to them on the screen yet. https://user-images.githubusercontent.com/59611559/74106383-86468900-4b5e-11ea-8d31-b308be920fcd.png https://user-images.githubusercontent.com/59611559/74106390-92324b00-4b5e-11ea-8bbc-3f7dd0753e5f.png I've also included a couple of lines of code (if statement) to deal with any files that may be like mux but without the CV columns. I will test this soon.

Will you be working on making the displayed RKA table(s) show only the top and bottom 10 kinases by log fold change, and on making the bar plot be sorted by log fold change?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/celiaccb/Software-Development-Group-Project-2020/issues/76?email_source=notifications&email_token=ANMWTDSZ2S46LW6OENAWWQLRCA2BJA5CNFSM4KR3QEXKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELGSESI#issuecomment-583869001, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANMWTDUOQBWPAVT3UO7FS2LRCA2BJANCNFSM4KR3QEXA.

[AnnaDearman]

It looked like I needed to add this, I haven't tested it yet:

[AnnaDearman]

I'll do the RKA table and bar plot thing then 😃

[East-YutangChen]

Cool. Good luck.

2020年2月9日 下午6:12，AnnaDearman notifications@github.com 写道：

I'll do the RKA table and bar plot thing then 😃

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/celiaccb/Software-Development-Group-Project-2020/issues/76?email_source=notifications&email_token=ANMWTDUFDDMBX5KGHDMEBMDRCBBSVA5CNFSM4KR3QEXKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELGTYCI#issuecomment-583875593, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANMWTDXMCAUCKUDXXW5EF3TRCBBSVANCNFSM4KR3QEXA.

[celiaccb]

Hey guys! Really impressed by the data analysis! I just changed application.py with the latest one, and changed datanalysis.html to support several inhibitors. However, I can't get it to show the KRA table for each inhibitor (it always show the KRA table for the first inhibitor only) I think maybe we need to save the KRA data frame for each inhibitor (df_submean2_20) with a different name, so like df{inh}_submean2_20 or something like that. I was just wondering if you guys know how to do that, otherwise I will try myself tomorrow.

[AnnaDearman]

Good point, I don't know why I didn't do that! Will do now.

[AnnaDearman]

I'm nearly there but I'm stuck. For some reason it's displaying every inhibitor's RKA table at the top, even though I've made a list of RKA tables and asked the datanalysis.html file to look at item [i] in the same loop as everything else! I can't remember exactly what else I tried but I did get it to place each "table" in the right place but it was a load of html, not a table.

[AnnaDearman]

I've put the datanalysis.html and application.py files on my fork that display the separate inhib RKA tables (but all together at the top), if anyone wants to have a look. I'm sure the answer is going to be very simple, but I'm going round in... a loop 😅

[celiaccb]

Since we managed to show the separate tables now, shd we say the data analysis part is done or do you guys want to change something?

[AnnaDearman]

We could close it because I did the bar plot sorting and the top 20 kinases thing as well, but does anyone want to comment on my comparison of the stats method against another website's method first?

[East-YutangChen]

Thanks guys, I don’t want to any more changes now.

2020年2月10日 下午3:11，celiaccb notifications@github.com 写道：

Since we managed to show the separate tables now, shd we say the data analysis part is done or do you guys want to change something?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/celiaccb/Software-Development-Group-Project-2020/issues/76?email_source=notifications&email_token=ANMWTDVT5FNBOVBS7N2OBMDRCFVCXA5CNFSM4KR3QEXKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELI3PUA#issuecomment-584169424, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANMWTDUJMB7SJAJ6TFYXZ6TRCFVCXANCNFSM4KR3QEXA.