Behavior of the mdl-createmodel command about model-seed-core HOT 3 CLOSED

Aaron, you should get the same result from using the web submission or the local Model SEED. Both systems use the same code. I apologize for the poor reliability of the web system. We had a number of problems with our queue throughout May and part of June. We have resolved these now, and we've put tests in place to ensure that we are notified immediately should the queue break down in the future.

Anyhow, you are using the right command. Sadly, the output to the screen is not saved anywhere. You can of course save this text yourself by piping it to a file using " > CreateModelOutput.txt" at the end of the mdl-createmodel command.

The biomass template does change depending on whether the organism is gram positive or gram negative, and the system does sometimes get the calls wrong. In these cases, the biomass reaction will have the wrong biomass. Fortunately, ModelSEED has a mechanism for over-riding the call and forcing the system to consider a particular genome as "gram negative" or "gram positive". Open this file: "Model-SEED-core/lib/ModelSEED/FIGMODELconfig.txt", and find the lines that start with: "%Gram positive" and "%Gram negative". Add your genome to the appropriate list. Make the same change to this file: "Model-SEED-core/config/FIGMODELconfig.txt", or just run a ms-config.pl. Note, you have to change both files to avoid overwriting your change when you run ms-config.pl.

Once you've made this change, you should be able to run the mdl-reconstruction command for your genome.

Note, I did need to make a small tweak in the code to make this work, so you should do a "git pull" before doing all this.

from model-seed-core.

Thanks for the workaround, Chris. While doing the code modifications, I noticed that there were lines for "families" where it looks like you are pattern matching on genus names. I modified that line to include the genus, since I'll be building many models for this genus.

This brings up another question though.. I tried building another model (so as not to overwrite the current model I have) and compare to the existing one using the mdl-createmodel command. This did not recognize that I had included the genus name in the families line. However, I followed that up by running the mdl-reconstruction on the newly generated model, and it did recognize the genus name and classified the organism correctly. Is this how this is supposed to behave? Does the code for the mdl-createmodel command need to be updated, as well?

from model-seed-core.

Aaron,

I just checked in the code so the "families" overrides work the same as the specific genome overrides, appropriately affecting the biomass reaction. This should give the behavior you want.

Chris

from model-seed-core.