Pheno
PlatformGenomic Reasoning

Know what it can make before you build it.

Routes, knockouts and inserts, read against your own genome.

A design that fails at the bench costs weeks. Pheno reads your genome as a map of what the organism can run, so the route, the knockout and the insert are settled on evidence before anyone picks up a pipette.

The map

See what your organism can actually run.

A map drawn from your genome, with direction and evidence on every reaction.

Host genes are matched to enzymes and enzymes to reactions, and the map keeps the evidence for each. Ask the same question under three direction policies and build on the routes that survive all of them.

Direction policies
PolicyABCDE
Accepted consensusonly reviewed direction evidence→→?→
Heuristic firstadds stated thermodynamic rules→→→→
Exploratory unionevery source, contradictions kept→⇄→⇄
Shared across policiesholdsvariesvariesholds
Directions come from reaction databases, curated enzyme records, stated thermodynamic rules and reviewed papers. Model priors only advise. When sources disagree the direction is unknown and the disagreement is kept.
Routes

Know which steps are yours and which are borrowed.

Host, reference or literature, labelled step by step.

Each step of a route is labelled host, reference or literature, and a gap the host cannot fill is shown with what could fill it. A search that ran out of budget says so.

Route search
donor step · EgtB/EgtC · reference onlyL-HistidineHercynineegt-1 · NCU04343Hercynyl-Cys sulfoxideegt-1 · NCU04343Ergothioneineegt-2 · NCU11365
Host steps
3 of 3 with gene support
Direction policy
Rhea + thermodynamic rule
Donor steps found
1, not compiled
Host route from the compiled organism topology, gene support by homology to a curated reference. The donor step is reference chemistry and stays out of the host graph.
Edits

Cut the drain. Bring in the gene.

Knockouts and inserts, ranked on evidence.

To raise a molecule, cut what consumes it. Pheno finds the drains that run in your organism and ranks donor organisms for the step you are missing, nearest relatives first.

Knockouts and inserts
targetdegradationside productcannot run hereknockout
Missing step: no host gene for this reaction
Donor organisms, nearest first
  1. Same species, reference strainclosest
  2. Same class of funginear
  3. Other fungifar
  4. Food-fermentation organismsflagged
A drain counts only if its reaction can run and the organism can reach it. A donor search ranks organisms by distance from the host; a gene absent from the reference strain means it is not annotated there, not that the activity is missing.
Your data

Read every answer against your own measurements.

Expression and metabolite pools beside each step.

See which genes switch on under each feedstock and how the target pool moves over time. A measurement is never dressed up as a yield.

Measured beside the route
AlmondSpentCarrotDistillers'OatOkaraSunflowerTomatoCNCNCNCNCNCNCNCN
step 1 gene23
step 2 gene323233
step 3 gene3
consumer gene-2-3-3-3
C: substrate as carbon sourceN: as nitrogen source down up
Target in the medium
extracellular pool, one substrate
0 h8 h24 h

Relative intensity. Comparable across samples of this metabolite only.

Expression is log2 fold change against a reference at one time point; it shows capacity, not flux or direction. Pools are extracellular and relative; detection is not production. Values here are generated.
Inputs

What it reads.

Every dataset is versioned on the way in, so a report can name the exact data it ran on.

Datasets
Genome
FASTA assembly with GFF or GTF annotation
Proteome
protein FASTA, indexed for local search
RNA-seq
differential expression per substrate and feeding mode
Metabolomics
per-sample intensities and time-course ratios
Reference chemistry
versioned reaction, enzyme and pathway databases
Media
named seed sets, or a list you supply
Papers
PDFs or identifiers; see Self-Evolving Knowledge