Do production, completeness and conservation signals indicate gene essentiality?

Several lines of evidence in the gene-essentiality synthesis treat what an organism can make, or what its genome encodes, as a guide to what it requires. Other evidence shows those signals do not settle the question. Metabolite production records cannot show uptake or dependency ₁. Many ortholog families (groups of genes descended from one ancestral gene) are not essential in every organism that carries them ₂. A pathway predicted complete in most mapped genomes can still be missing in one ₃. This matters because the synthesis ranks direct perturbation above predictors, and reading capability signals as essentiality calls would promote predictors beyond what they can carry.

Evidence Sides

Side A: capability and conservation signals are widely shared and look like a common core

• Of 17,222 ortholog families, 859 were universally essential, meaning essential in every organism where the family has members ₂. This is a shared essential core among member organisms, not proof of presence or essentiality across the whole sample.
• By pathway-completeness prediction, 17 of 18 amino-acid pathways were complete in all 7 mapped organisms ₃.
• Web of Microbes (WoM), an exometabolomics database recording metabolites organisms release into growth media, supplies production matches linked to organisms’ metabolic output ₁.

Side B: these signals cannot establish uptake, pathway completeness, or essentiality

• WoM production matches cannot establish uptake, pathway completeness, or gene essentiality ₁.
  • The snapshot has no organism consumption actions ₁.
  • 107 formula-only matches (matches on molecular formula alone, without structural identity) expanded to 900 candidate molecules ₁.
• Most ortholog families are not universally essential: of the 17,222 families, 4,799 were variably essential and 11,564 were never essential ₂.
• Uniform completeness has an exception: Desulfovibrio vulgaris (DvH) lacked predicted serine biosynthesis ₃.
• A genomically complete pathway can show no detectable fitness importance under the conditions tested, so completeness does not by itself indicate dependency ₄.

Possible Reconciliations

• Hypothesis 1: Capability, conservation and dependency are distinct layers. Production records, completeness predictions and family conservation describe what organisms make or encode; essentiality would require perturbation evidence under defined conditions. Under this reading, Side A’s signals are real but not interchangeable with essentiality.
• Hypothesis 2: The DvH serine gap may be a prediction artifact or a genuine auxotrophy (a requirement for an externally supplied nutrient). The result is a pathway prediction, not a growth measurement ₃.
• Hypothesis 3: Only the universally essential families may track essentiality within the conserved signal, with variably essential families reflecting genomic context rather than capability alone.

Resolving Work

• WoM and transposon fitness: Join WoM production records to transposon-sequencing (TnSeq, which infers gene importance from mutant abundance) fitness data for shared organisms, testing whether genes behind emerged metabolites are fitness-relevant.
• Formula-only matches: Re-score the 107 formula-only WoM matches against structure-resolved standards or spectral libraries to ask how many of the 900 candidates are real ₁.
• DvH serine: Grow DvH in serine-free versus serine-supplemented media to test whether the predicted serine gap is a true auxotrophy ₃.
• Variable essentiality: For the 4,799 variably essential families ₂, model essential status against pathway completeness and gene-neighbourhood context, asking whether context explains variability better than conservation.
• Completeness versus dependency: Cross-tabulate the 17 complete amino-acid pathways in the 7 mapped organisms ₃ against per-gene TnSeq fitness, asking whether completeness predicts dependency within this sample.