Proportional hazard analysis for survival during paraquat exposure and starvation.

Proportional hazard modeled for ch+/- and d4eBP as single and double mutants relative to wildtype, and ch-/- and d4eBP as single and double mutants, with likelihood ratio test. Coefficient β for single loci: when significantly less than zero indicates reduction in mortality, estimates greater than zero indicate elevated mortality. Double mutants: coefficient β when significantly different from zero indicates gene interaction where the effect of the double mutant differs from expectation from product of single mutants. Epistasis inferred when significant gene interaction increases mortality (positive β) relative to expected product of chico and d4eBP; synergy inferred when gene interaction reduces mortality (negative β) relative to product of chico and d4eBP genotypes. Female and male survivorship when exposed to paraquat. Female and male survivorship during fasting.