It says that some genes result in the same outcome when knocked out as other genes, and identifies novel genes that putatively participate in the same pathways as others. This helps get at the potential function of genes without known functions.
So, people do this a lot and frequently make mistakes. For example, when you knock out a gene, you also damage any overlapping genes (yes, genes can overlap). most studies don't pay attention to the damage they do to overlapping genes.
The underlying physical model for how gene products interact to make phenotypes ends up being so hopelessly complex and latent that most conclusions in this area end up being "sufficient, but not necessary" instead of "necessary, but not sufficient"
> It says that some genes result in the same outcome when knocked out as other genes
a very useful map to make, but I don't see that this contradicts my comment - both genes in this case are dependencies to the outcome, without either of them, the outcome fails
this does not sound to me like we know the "function" of these genes, only that they're nessary for each phenotype
not to knock the research, just trying to make sense of what they're really mapping, in my own language of computer code (i suppose "function" has a different connotation in genetics)
> this does not sound to me like we know the "function" of these genes, only that they're nessary for each phenotype
But they're not even measuring the phenotype. They're using the transciptional signature as a substitute for phenotype/cell function (i.e. the bag of RNA model). This is a poor substitute if you try to apply this to practical applications such as cell engineering. Let's say I perturb a cell to match it's transcriptional signature to that of a neuron. Does that make it a neuron? Not if it doesn't function like a neuron.
I think this is a really important point. With a few exceptions (like the neat implied aneuploidy assay), they haven't measured an outcome or phenotype for the genes. They have measured the impact on transcription (well, mRNA levels, via transcription or some other effect). That is an extremely useful dataset, but it's not enough to say what the phenotypic effect of knocking out any given gene is, much less what the actual mechanistic function of the gene product is.
It's also important to note that there are loads of genes whose effects are not mediated by changes in mRNA levels. If you knock out Arp2, a cell can't move properly, because Arp2 is involved in assembling cytoskeletal structures needed to do that, but you probably won't be able to tell that by looking at the cell's mRNA.
> "this outcome fails when this byte of data is missing"
The outcomes here are not failures, they are measurable phenotypic differences, which they use to group genes into phenotypic outcomes. The typical "knockout -> failure to perform a function" is not what's being measured here.
It says that some genes result in the same outcome when knocked out as other genes, and identifies novel genes that putatively participate in the same pathways as others. This helps get at the potential function of genes without known functions.