Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Taneli Pusa
    @htpusa
    Hello everyone. Is there a way to change the font size on the reactions/metabolite labels?
    Panache123
    @Panache123
    @zakandrewking Hello! I am new to Escher, so I have some problems in creating my own map. What I have done is: 1. I have my own cobra sbml model named 'Bacteroides_caccae_ATCC_43185'. I want to make a map to show all reactions and metabolites of this model. 2. I have changed its type to .josn. But after loading it in escher website, it shows nothing. I suppose I missed some steps to make the map. Could you please help me out?
    Zachary A. King
    @zakandrewking
    hi @Panache123 let's discuss in the DM
    j-robs
    @j-robs
    Hi @zakandrewking , when using Escher I've noticed an issue (or feature) where if I want to drag a reaction somewhere on the canvas all the control points on the lines stay in the exact place where they were, so the connecting lines end up all over the place and have to be dragged individually. Is there any way this can be made so that you can select them with a selection box when they're showing so that they move with the reaction like they do when you use the rotate tool? Sorry, I'm not very familiar with GitHub and relatively new to coding so I couldn't find where this is set in the code to be more helpful. If there's already a way to do this in the program it would be great to hear! Thank you.
    mjkijsc4
    @mjkijsc4
    Dear anyone who can help? I'm wanting to use ESCHER to visualise fluxes. I've read the documentation, and have to be honest,i'm struggling. I have my SBML model in json format and now wanting to build a map of key pathways. Can anyone help?
    Zachary A. King
    @zakandrewking
    @j-robs I made a gif to see if we're on the same page here: http://g.recordit.co/HEUx8CmhIB.gif This is what you have now? And it's annoying to move each of the bezier points after the drag like I do in that gif?
    @mjkijsc4 can you describe where you are stuck right now? feel free to send me a DM if you want to chat (hover over my username and click Chat Privately)
    j-robs
    @j-robs
    @zakandrewking Yes that's it. Weirdly though when I click and drag select like you did there none of the points move with the drag at all. I can see in yours that some of them moved with it but when I move a reaction say to the other end of the canvas all of those points remain in the original place, like this:
    https://gyazo.com/cfabbf69f1599175654d31aa461acc1a
    Björn Johansson
    @BjornFJohansson

    I used cobrapy to calculate the maximum ehtanol yield from glucose using the iMM904 yeast model.
    The results were correct, but the reactions do not map to the escher map in some cases. See cell 19 here:

    https://nbviewer.jupyter.org/github/BjornFJohansson/cobrapy_exercises/blob/master/cobrapy_ethanol_yield_iM904.ipynb

    A good example is the HEX1 reaction on the canvas which is zero since the model
    uses the GLUK reaction for what is essentially the same reaction. Do I have to delete these reactions in the model
    or can I make escher show the correct reaction?

    Zachary A. King
    @zakandrewking
    @j-robs that's not good. Do you get it all the time? On the Escher website or local installation?
    @BjornFJohansson yes, you can edit the map to remove add the GLUK reaction, but you'll have to do it on the Escher website right now (https://escher.github.io/). Load you map and your model in the dropdown menus, and change it to have the reaction you want, then download the map JSON file for use in the Jupyter notebook (we are working on a new version where you can do all of this in the notebook, but it's not out yet)(
    Björn Johansson
    @BjornFJohansson
    Thanks @zakandrewking Ill give it a try.
    j-robs
    @j-robs
    @zakandrewking I do get it all the time on the escher website on multiple computers and colleagues have experienced the same issue. I'll try the local installation now.
    Zachary A. King
    @zakandrewking
    Any idea how to recreate it from a fresh load of the page?
    j-robs
    @j-robs
    Right so it seems like this issue only happens when I load in reactions I create myself. I'm making a script for quickly loading reaction pathways into Escher from a metabolic network and the issue seems to occur when I load in my reactions whereas when I use the default add reaction tool in escher the nodes move fine. Maybe an issue with how I'm defining segments? I'm setting b1 and b2 as null when I load them in because that seemed to be the default in your json files is that right?
    Zachary A. King
    @zakandrewking
    I’d have to take a look. Can you share an example of a file that’s causing the problem?
    Zachary A. King
    @zakandrewking
    hey, i found the issue!
    if you run your json maps through a jsonschema validator (e.g. here https://www.jsonschemavalidator.net/) against the schema from our github repo (https://raw.githubusercontent.com/zakandrewking/escher/master/jsonschema/1-0-0), you'll see errors pop up
    like this: Message: Invalid type. Expected String but got Integer. Schema path: https://escher.github.io/escher/jsonschema/1-0-0#/items/1/properties/reactions/patternProperties/[0-9]+/properties/segments/patternProperties/[0-9]+/patternProperties/(from|to)_node_id/type
    i.e. from_node_id and to_node_id should be strings, not integers
    we don't do jsonschema validation when loading escher maps in the website to avoid the performance hit, although that might be smart to do in the future
    in the meantime, if you change those values to strings, things will start working as expected
    Zachary A. King
    @zakandrewking
    @j-robs ^^ that was to you
    j-robs
    @j-robs
    @zakandrewking thank you that's great, all working fine now!
    Ayush Mandwal
    @MandwalAyush_twitter
    @zakandrewking Hi, I am trying to use time-series data visualisation using Escher. I have downloaded the files from Github but having some troubles installing the files. Can you help me out in how to use this code for visualisation time-series data? Here is the link to Github file- https://github.com/christophblessing/escher
    Zachary A. King
    @zakandrewking
    @MandwalAyush_twitter That feature is not finished yet, so I can't help you with this at the moment. We do have plans to finish it after the release of Escher 1.7, so keep an eye on our escher releases for updates: https://github.com/zakandrewking/escher/releases
    Ayush Mandwal
    @MandwalAyush_twitter
    But the paper- "Visualizing metabolic network dynamics through
    time-series metabolomics data" claims to use Escher to show time-series data in a animated way. Are you saying that the code used for getting such visualization is not ready yet? I was thinking of using it for my work. I am also trying to visualise my metabolomic data with same idea. And if you are updating the package, how long will it take for you to get it done? I am excited to use it for my own work.
    Ayush Mandwal
    @MandwalAyush_twitter
    Also I managed to install Escher from setup.py file from the file I downloaded from Github. The problem was that package.json file was missing in escher folder. When I copied it in escher folder, I was able to install it.
    Zachary A. King
    @zakandrewking
    Yeah, we did use that development branch to generate the results from the paper. We just haven't been able to wrap it up for general use yet. But feel free to try it out in the current state (that's why we love open source :). And let me know if I can help with anything
    Ayush Mandwal
    @MandwalAyush_twitter
    That's great. I need help in that only. I have looked at different python files but they all have just function definitions. How do I use any of them to get the results as shown in the paper? Can you share if there is any workable sample code which illustrate those features and allow me to edit it so that I get my results displayed as in the paper. Thank you so much...
    We can start separate chat if you prefer...
    Tomek Diederen
    @Patrickens
    Hey, is there any way of automatically generating escher maps from all reactions that carry non-zero flux (always less than 200 reactions in my case) for the iJO1366 model?
    and if not, what is the easeiest way to do this? I have no clue how to draw coordinates properly or how to format the JSON.
    Tomek Diederen
    @Patrickens
    And is there a place with some examples of using the python API where reaction_data,metabolite_data and gene_data are used?
    Jeremy Zucker
    @djinnome
    @Patrickens I don't think it would be easy to do that with escher, but d3flux can probably do what you want: https://github.com/pstjohn/d3flux
    Zachary A. King
    @zakandrewking
    @Patrickens definitely try out d3flux. I'd be curious to hear how it works for your need there. In Escher, you can't draw automatically, but if you load flux reaction data in the app, it will suggest the highest flux reactions so you can draw pathways more easily.
    Ayush Mandwal
    @MandwalAyush_twitter
    Why the current active version of Escher software does not export images of the network? It shows the options. Can one export image or not?
    Zachary A. King
    @zakandrewking
    Can you share a little more detail on the bug? Ie Which button did not work? What browser and operating system are you using?
    Ayush Mandwal
    @MandwalAyush_twitter
    I was trying on chrome version-74 and mozilla version-66. I tried to both exporting png as well as svg format but do not work.
    I can save map in json format though.
    Ayush Mandwal
    @MandwalAyush_twitter
    Also the latest version of Escher where we can simulate data does not allow changes in color or size of nodes when we add metabolic data. Can you help me in that? The color and size are all set to some default values. I need to change them for the analysis.
    Ayush Mandwal
    @MandwalAyush_twitter
    i am using ubuntu as operating system
    Zachary A. King
    @zakandrewking
    thanks for the bug reports. i see the saving issue – will fix it ASAP. I don't see the problem with metabolomics data sizing & coloring. any chance you can share the data file you used? or describe the steps you followed to see the problem?
    Zachary A. King
    @zakandrewking
    FYI map saving should be fixed now
    Ayush Mandwal
    @MandwalAyush_twitter
    Thank you. For metabolomics data, resizing has issues with the latest version of Escher which hasn't released yet. Current active version of Escher has no problem with resizing nodes.
    Zachary A. King
    @zakandrewking
    Ok thanks
    Nikolaus Sonnenschein
    @phantomas1234
    I stumbled over a map that has an additional notes property for genes referring to original_bigg_ids. escher.github.io doesn't complain about it even though it is not a valid map according to https://github.com/zakandrewking/escher/blob/master/jsonschema/1-0-0. Then I figured that quite a lot of these maps exist on GitHub https://github.com/search?q=original_bigg_ids&type=Code Does anyone know how this came about? Is this maybe the result of some kind of automatic conversion? Should the schema be relaxed maybe to to allow additional properties at https://github.com/zakandrewking/escher/blob/master/jsonschema/1-0-0#L96?
    Zachary A. King
    @zakandrewking
    not sure how those got in. But most of the results from GitHub are COBRA model JSON. Escher ignores extra fields, but I'd recommend that people drop those extra fields if they are using the maps a lot (e.g. developing extensions). I haven't gotten any requests for a "user_data" field in the Escher schema, but we could add that if desired by users.