-
Notifications
You must be signed in to change notification settings - Fork 44
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Human 1.17 #700
Human 1.17 #700
Conversation
… but remove the FAD-dependent version of the SDH reaction
…thesis (even-chain)
…ich I apparently forgot to do earlier
…pt reactions in model/reactions.tsv
- This commit includes fixing typos, and reverting back to original way of extracting subsystem values
…_synth Remove Mitochondrial Bile Acid Synthesis Pathway
fix: update name for GAP in CellfieConsensus metabolic tasks
remove rows 'MAR01639', 'MAR01635', 'MAR01637', 'MAR01653', 'MAR01656' from "reactions.tsv"
Fix FAD/Oxygen-Peroxide Duplicate Reactions
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Wow, a lot of changes in this update - great to see the ongoing progress and improvements!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Very nice!
@haowang-bioinfo will we see the essentiality results, like last time?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
yes, indeed! |
updated Hart2015 essentiality results:
|
Updated essentiality evaluation using combined (
The trend of MCC is not very promising |
Indeed. How about we pause this release and investigate deeper? |
It looks like the number of false positives is increasing quite a bit. This could be the result of e.g., removal of duplicate pathways, removal of incorrect isozymes, or closing of some infeasible loops - all of which are generally good things. So I agree with @mihai-sysbio that it would be worth investigating, but I also wouldn't be too alarmed. |
it's a good idea to investigate further for the MCC declining, any attempt is welcome. while the pause of this release doesn't make sense, because:
|
For the sake of the debate, I'm going to argue for the pause. Don't get me wrong, I am not entirely convinced pausing is the best way forward. I am just willing to engage in the debate. A pause is not a cancellation, it's a delay, and the model can still be used. The model is openly available, even though not in the
Moreover, now the information is still relatively fresh, so digging into this now might be more comfortable than in say in 6 months. And what we will learn from this process will also be benefiting the next release. |
Some more thoughts: The way I've mapped the MCC score in my head is like this: the interval [-1, 0] is "worse than useless," and the interval [0, 1] corresponds to 0% to 100% usefulness. With this perspective, a reduction of 5% in usefulness from 21% usefulness is not good. Of course, this thinking really abuses the numbers. Another thing: if a new release would be affecting the MCC by say 0.2 (instead of 0.05 like now) we would likely take it very seriously. So then the question is not if but how much of a MCC change is acceptable with a release. |
Debate is good, it often makes things clearer. Probably my previous message is a bit ambiguous, so try again. My key point is investigating MCC and making this release can be separated, and proceed in parallel. They are not conflicting with each other, no need to pause one for another.
yes of course, I totally agree to check this out. Actually this and all previous releases, as well as all involving transparent changes, enable this kind of check.
Agree, literature evidence, MCC, and some other indicators may still not be sufficient. But this shouldn't stop rational changes with critical review.
yes, just go ahead please
I'm not exactly sure what cutoff should be used for evaluating MCC, which is only one dimension of assessment. So any follow-up analysis in clarifying this is welcome |
Main improvements in this PR:
geneShortNames
field (introduced by add gene name to Human-GEM #539) by padding with empty elements in functionaddMetabolicNetwork
updateAnimalGEM
with option of reusing Human-GEM biomass equationMAM01268
orMAM02757
in all compartments, and 10 metabolites and 22 genes that only participated in these reactions, as discussed in the comments on Pairs of reactions with identical stoichiometry but different GPRs #580rxnNotes
PMID:19357427
as referenceMAM02630x + MAM01803x -> MAM02041x + MAM01802x
, GPR:ENSG00000161533 or ENSG00000168306 or ENSG00000087008 or ENSG00000110887 or ENSG00000203797 or ENSG00000007171 or ENSG00000148832 or ENSG00000179761 or ENSG00000158125
, with reference:PMID:30378035
reactions.tsv
, as a tentative implementation according to the discussion in How to better document deprecated met/rxn ids #615reactions.tsv
reactions.tsv
ENSG00000017483 or ENSG00000188338
simplifyGrRules
function to fix compatibility with Matlab "children" function, which has been changed since 2020b, to resolve issue described in update use of functionchildren
insimplifyGrRules
#493