This page records a tension from cofitness-network-architecture: the functional-dark-matter work presents co-fitness as a route to function for genes of unknown function, but its own coverage and scope statements limit where that route applies. Co-fitness means correlated fitness profiles across conditions in RB-TnSeq (random barcode transposon sequencing) experiments. The disagreement matters because it decides whether co-fitness evidence can be used for essential dark genes and for lineages outside the Fitness Browser’s taxonomic sample. 57,011 dark genes were found 1, but 32,075 co-fitness-tested pairs are described in two incompatible ways 1.
Evidence Sides
Side 1: co-fitness as a broad inference route, including essential genes
The functional-dark-matter analysis found 57,011 dark genes 1. One section of that report describes its 32,075 co-fitness-tested pairs as providing the strongest functional inference for essential genes that lack direct fitness data 1.
Side 2: co-fitness restricted in scope and in coverage
Finding 12 of the same report specifies the 32,075 pairs as non-essential operon pairs (an operon being a set of adjacent genes transcribed as one unit) and states that co-fitness is unavailable for essential genes 1. The tested-pair scope and the claimed applicability therefore conflict 1.
Coverage is also skewed. The Fitness Browser contained 37/48 Pseudomonadota and no Archaea, Actinobacteria, or Epsilonproteobacteria among the top 500 candidates 1.
The truly-dark analysis adds a 17,479-gene pangenome-linkage gap. A pangenome is the full gene set across genomes of a species. The analysis estimates approximately 2,841 additional truly dark genes among those 17,479 at the 16.3% linked rate. This is an estimate, not an observed count 2.
Possible Reconciliations
- Hypothesis A: “strongest functional inference for essential genes” may refer to operon-neighbour inference, with co-fitness of non-essential partners used only as indirect support. If so, the Finding 12 scope is correct and the other section overstates applicability. Neither reading is established by the TENSION text.
- Hypothesis B: the coverage skew may reflect only which organisms the Fitness Browser sampled, not where dark-gene biology is concentrated. Under this hypothesis, conclusions would remain untested outside Pseudomonadota, and their validity within Pseudomonadota would still need its own test.
- Hypothesis C: the 16.3% linked rate may or may not transfer to the 17,479 unlinked genes. Until those genes are assessed, the approximately 2,841 figure remains an estimate, not an observed count 2.
Resolving Work
- Pair-level audit: take the 32,075 tested pairs from the functional-dark-matter outputs, join each member to essentiality calls, and ask whether any pair contains an essential gene. This would establish the tested-pair scope, though not whether indirect inference reaches essential genes.
- Unlinked-gene reannotation: reannotate the 17,479 unlinked dark genes directly, without a pangenome link. Ask whether the observed truly dark count matches the approximately 2,841 estimate.
- Eligible-pool audit: tabulate, by phylum, the full dark-gene pool and organism coverage from which the top 500 candidates were drawn. Ask whether the 37/48 Pseudomonadota share tracks the sampled organism composition or the ranking criteria.
- Gap-filling lineages: add fitness data for Archaea, Actinobacteria, or Epsilonproteobacteria. Then test whether co-fitness partners of dark genes behave as they do in Pseudomonadota.