starplast.drug_sensitivity

Drug sensitivity per gene: does disrupting it change survival under a compound.

The second curated table in the map, and it exists for the same reason resistance does -- no genome-wide chemogenomic screen has been published for Toxoplasma, so the answer lives in individual knockout studies rather than in a deposit. Six catalogue sweeps looked for the screen and correctly did not find one; what they were testing was the sentence "no screen exists", which is not the same statement as "the question cannot be answered". A knockout with a measured shift in sensitivity to a named compound answers it one gene at a time.

The bar

A row exists only if a knockout or knockdown showed a measured change in sensitivity to a named compound, read from primary open-access text. Not "an inhibitor of this protein kills the parasite" -- that is target engagement and has its own slot. The direction is recorded, and so is unchanged, because a transporter deleted with no effect on analog sensitivity is a result and suppressing it would leave the column looking like a list of hits.

What the background qualifier is for

TgENT3's effect was measured in a ΔTgAT1 background rather than alone, so it is a genetic-interaction result and not a clean single-gene one. background says so per row. Reading those two rows as though the gene had been deleted on its own would be the same error as reading a double mutant's phenotype off one of its genes.