Closed AndrewCRMartin closed 2 years ago
This should have been fixed now. There are no longer blank domain comments in the expression.
Almost....
As before, if I create an IgG, add two X domains that are too far to interacts and then set them as TYPE:ZIPPER
all seems OK except that after doing Tidy
they are labelled as MOD:ZIPPER
instead of TYPE:ZIPPER
Have a try of it now. Hopefully the comments system has been fixed now.
Still not quite working right.
After a little more playing, conclusions are
It's good to see that I can add something like MOD:STABILIZATION to hinges (as that is very common with IgG4 drugs)
I think I see what's going on. Since the Linkers are drawn to the centre of the X domains then if you put the comment where the X domain and the Linker overlap (i.e. towards the top of the X domain, then it thinks the comment also applies to the linker. I'll work on a fix for this but placing comments where the linkers do not overlap just results in the X domain being assigned the comment.
There was also an issue in the comment system I wrote yesterday which split multiple comments up, leaving you with just the first comment, which has now been fixed.
It was having trouble recognising X* domains so I've removed those for now, but they can be written back in.
Is this the result that you wanted (let me know if you want X* domains to replace the X domains)?
Not sure that was the problem I was seeing - it was the HINGE that was getting the MOD:OTHER, not the linker!
Surely an X is exactly the same as X but has a next to the X in the label?
On Fri, 11 Feb 2022 at 09:39, James Sweet-Jones @.***> wrote:
I think I see what's going on. Since the Linkers are drawn to the centre of the X domains then if you put the comment where the X domain and the Linker overlap (i.e. towards the top of the X domain, then it thinks the comment also applies to the linker. I'll work on a fix for this but placing comments where the linkers do not overlap just results in the X domain being assigned the comment.
There was also an issue in the comment system I wrote yesterday which split multiple comments up, leaving you with just the first comment, which has now been fixed.
It was having trouble recognising X* domains so I've removed those for now, but they can be written back in.
Is this the result that you wanted (let me know if you want X* domains to replace the X domains)?
[image: Screenshot 2022-02-11 at 09 36 34] https://user-images.githubusercontent.com/73530578/153568426-926210ae-c3c1-42c0-8521-0ddfac6652f3.png
— Reply to this email directly, view it on GitHub https://github.com/JamesSweetJones/abYdraw/issues/1#issuecomment-1036024896, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABBQ7UMGNAKX6VIZ2WAL6FDU2TKM3ANCNFSM5N6ITRNA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.
You are receiving this because you authored the thread.Message ID: @.***>
-- Andrew C.R. Martin Professor of Bioinformatics and Computational Biology. Institute of Structural and Molecular Biology, UCL http://www.bioinf.org.uk/ Twitter: @PrfAndrewMartin
Try again now. I've made some changes now so hopefully the hinges should not randomly pick up modification comments.
Seems to resolve the problems above now, BUT if I add TYPE and MOD comments to the X domains and then try to add a *
to them the comments get removed.
I know you said you had removed it, but when I try to add the *
, it renders on the image, but then on doing a Tidy
it isn't in the AbML (and therefore disappears from the display) - as well as resulting in the comments going!
To remove comments, you can just delete them from the canvas, and then by clicking Tidy
the comment will be removed from AbML. I can work out something for the X domains to fix this.
That wasn't quite my point.
If I add MOD/TYPE to an X domain, then try to add a *
, the MOD/TYPE is removed.
Tidy
All is fine at this stage
*
X
domainTidy
The MOD:
annotation has gone.
OK, I may have come up with a solution to this. Try again now.
Seems fine now! Will do some more testing then close
Did this work, Andrew?
I'm now getting problems with the linker getting the TYPE: label.
X
and add an X domain at the bottom of each chain - not close enough to interactLinker
and add linkers to the new X domainsTidy
Type
Type
pulldown to select (e.g. TYPE:ZIPPER)Tidy
Both the X domains and the linkers have the TYPE: info
Tidy
againNow the linker annotation disappears!
The other problem with adding a * to X domains has also come back!
If I add TYPE to an X domain, then try to add a *, the TYPE is removed.
All OK to that point. Now
The TYPE: annotation has gone.
I think, but haven't checked, that if you just click one of the X domains, the TYPE annotation goes from just that domain.
I was under the impression that a "" symbol referred to a modification and so I have only asked it to render when a "MOD:" label has been placed on a domain. Am I correct here or should "" be on the domain whenever a comment TYPE:/LENGTH:/NOTE:/ANTI:?
The *
is indeed only for a modification, but it is independent of the comments. i.e. you can have a *
without a MOD:
. So, if you add a *
it should be rendered whatever comments are (or are not) present.
The problem here was that adding the *
removed the TYPE:
comment.
Ideally a *
would be added automatically if you add a MOD:
comment, but that's not necessary.
Yes the reason for this is because the TYPE comment is linked to an "X" domain, but by adding a "" symbol to the domains it makes it an "X" domain, so when you run get AbML
it doesn't find the "X" domain the comment is paired to so it disappears. I will change it so these run independently.
I've made a fix for this. When adding/removing the "*" mod from the X domains, the TYPE labels no longer disappear
Just tried it with the Windows version and I still have a problem with labels going:
Tidy
*
Tidy
The *
is present, but the MOD: comment has gone.
Actually, just clicking Tidy
twice removes the labels:
Tidy
(the comment moves to the bottom as it should)Tidy
(the comment disappears)Hi Andrew, I've made another fix that seems to work when running through the scenarios you've outlined above. Let me know if you still have these issues. I have now set it up so "*" are automatically added to the AbML when a MOD label is placed on that domain.
I must also say that if a domain has "" on it but does not have a MOD comment, then by clicking Tidy
the "" will be removed. In this way, users will be made to give a comment, even if it is MOD:OTHER
. Do you think this is suitable or would it be best if we allowed "*" without having to explain it with a comment?
The specification doesn't require a '' to have a MOD comment. Indeed I'm using '' in the INN annotations since the information about modifications is given in more detail elsewhere. Comments are always optional.
A.
On Wed, 16 Feb 2022 at 16:10, James Sweet-Jones @.***> wrote:
I must also say that if a domain has "" on it but does not have a MOD comment, then by clicking Tidy the "" will be removed. In this way, users will be made to give a comment, even if it is MOD:OTHER. Do you think this is suitable or would it be best if we allowed "*" without having to explain it with a comment?
— Reply to this email directly, view it on GitHub https://github.com/JamesSweetJones/abYdraw/issues/1#issuecomment-1041829671, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABBQ7UIQXPKJNKOFVZ357K3U3PEA3ANCNFSM5N6ITRNA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.
You are receiving this because you authored the thread.Message ID: @.***>
-- Andrew C.R. Martin Professor of Bioinformatics and Computational Biology. Institute of Structural and Molecular Biology, UCL http://www.bioinf.org.uk/ Twitter: @PrfAndrewMartin
This new behaviour of not allowing a *
without a MOD:
breaks some of the library structures - e.g.
Orthogonal Fab
Tidy
The AbML has changed.
IMHO, if you input some AbML, click Get Structure
and then click Tidy
without doing anything else, the AbML should never change!
My issue here is that I use the *
symbols in the Get AbML
pipeline to identify which domain the comments apply to. Otherwise if there were more of that type of domain, the pipeline would assign the comment to all domains of that type. If this strategy isn't working I will come up with something better. I have a few ideas that could overcome this.
Each domain has a number so surely the comment should be attached via the domain number rather than the domain label.
I don't quite understand the issue though as any domain with a MOD comment will always have a '', but you can have a '' without MOD comment.
On Thu, 17 Feb 2022 at 12:10, James Sweet-Jones @.***> wrote:
My issue here is that I use the * symbols in the Get AbML pipeline to identify which domain the comments apply to. Otherwise if there were more of that type of domain, the pipeline would assign the comment to all domains of that type. If this strategy isn't working I will come up with something better. I have a few ideas that could overcome this.
— Reply to this email directly, view it on GitHub https://github.com/JamesSweetJones/abYdraw/issues/1#issuecomment-1042884839, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABBQ7UIUGAK26PYU4CQES4LU3TQTBANCNFSM5N6ITRNA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.
You are receiving this because you authored the thread.Message ID: @.***>
-- Andrew C.R. Martin Professor of Bioinformatics and Computational Biology. Institute of Structural and Molecular Biology, UCL http://www.bioinf.org.uk/ Twitter: @PrfAndrewMartin
I've rewritten the system that recognises and allocates comments. I think this is working much better and after having tested it on the scenarios outlined above!
Great - I will take a look :)
A.
On Thu, 17 Feb 2022 at 16:23, James Sweet-Jones @.***> wrote:
I've rewritten the system that recognises and allocates comments. I think this is working much better and after having tested it on the scenarios outlined above!
— Reply to this email directly, view it on GitHub https://github.com/JamesSweetJones/abYdraw/issues/1#issuecomment-1043147807, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABBQ7UKSPUJ4MF4UHX5C67LU3UOHNANCNFSM5N6ITRNA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.
You are receiving this because you authored the thread.Message ID: @.***>
-- Andrew C.R. Martin Professor of Bioinformatics and Computational Biology. Institute of Structural and Molecular Biology, UCL http://www.bioinf.org.uk/ Twitter: @PrfAndrewMartin
Much better, but still not quite there!
I created an IgG and tried clicking on each pair of interacting domains in turn and clicking Tidy
after each pair. It all worked fine until I did CL/CH1 (which I did last) when it gave:
Exception in Tkinter callback
Traceback (most recent call last):
File "/usr/lib64/python3.7/tkinter/__init__.py", line 1705, in __call__
return self.func(*args)
File "./abYdraw.py", line 7658, in <lambda>
button = tk.Button(frame, text = "Tidy", bg = "#CFCFCF", font=20, command=lambda: sequence_render_pipeline(lower_canvas))
File "./abYdraw.py", line 3810, in sequence_render_pipeline
exp = sequence_pipeline(lower_canvas)
File "./abYdraw.py", line 4399, in sequence_pipeline
bondx2 = bonds_dict.get(full_chain[-1])[0][2]
TypeError: 'NoneType' object is not subscriptable
Exception in Tkinter callback
Traceback (most recent call last):
File "/usr/lib64/python3.7/tkinter/__init__.py", line 1705, in __call__
return self.func(*args)
File "./abYdraw.py", line 5686, in items_selected
index = i[0]
IndexError: tuple index out of range
Also if I just do
IgG
*
CL
and CH1
(on one or both arms)Tidy
The AbML is created correctly, but the *
on CL doesn't appear
Tidy
again and the CL* changes to CL in the AbMLSorry!!!!!
Oh that is strange. Must be where I do my checks to make sure the labels aren't duplicated. I'll have a look.
Works now :)
Brilliant :)
This renders correctly with a TYPE:ZIPPER label on each X domain.
Now click 'TIDY'
If the domains are close enough to interact then the resulting AbML has blank MOD: comments for the X domains instead of the TYPE: comments. If they are not, then both the X domains and the Linkers have blank MOD: comments.
In either case, the rendering step doesn't work and I get the error message: