PROM results

[AgustinPardo]

Hello,

I am experimenting with PROM code of Chandrasekaran 2010. I am using a COBRA model (.mat) from Palsson 2018 (https://bmcsystbiol.biomedcentral.com/articles/10.1186/s12918-018-0557-y), also it is in the BIGG database in .json format, so I think it is well-formatted.

I could run the PROM in Matlab but I think I still don't understand the result of the script. It returns two variables, "f" and "v". (As PROM documentation said: the algorithm gives the growth rate (f) and flux response (v) after knock out of all the regulators in the regulatory model).

But I don't understand what those arrays (f ,v ) mean because they don't let you know which genes were knocked out, and also the dimensions don't fit my inputs.

Do you have any information that can clarify the results of the PROM approach?

Regards

Agustin Pardo

[jseidel5]

Hey,

there could be a problem when using newer metabolic models with prom as they use updated conventions and prom uses older ones. At least I had problems with that when trying to reproduce the original results. Prom should knockout all regulators in the order defined in the table with the regulatory interactions. There is also a Promv2 version attached to a paper about a method called GEMINI by the same authors. This version is also in the 'original' directory in this repository.

[AgustinPardo]

I also run PROM with the test data (data of the paper) and the result is the same. It returns two vectors, "v" and "f", with strange dimensions.

The "f" is for knocked out genes. But it is a vector of flows, so I still don't know which genes they are. But it has the dimensions of the number of regulators, so it makes some sense.

The "v" still causes me doubts because it does not map the rxns (reaction structure of the model) or any other structure of the model or inputs.

Do you have insights about this?