Fig. 7

Ppa-nhr-10 mutations disturb the expression levels of 37 putative regulatory targets in starving P. pacificus. Bold names along the y-axis correspond to names of the misregulated putative targets (El Paco v3) which are likely to cause the observed starvation-survival phenotype. Genes which have an exact C. elegans ortholog were additionally given the public gene name of this ortholog (indicated in brackets). Panels indicate whether these targets are members of the SRGs or not. Labels within the panels indicate functional information for the gene in form of WormCat terms (black) and/or predicted PFAM domains (gray). WormCat annotations were adapted from C. elegans, based on Cel-Ppa-orthogroups identified with OrthoFinder. PFAM domains were only indicated for genes we could not assign a WormCat term to, or genes whose WormCat terms ended in “unassigned”. Cutoffs used in the differential gene expression analysis with DESeq2 and for PFAM domain prediction with HMMER are indicated in the lower left. Note that, in the current El Paco v3 annotations, PPA25752 (in SRG targets) seems to be a gene model that artificially fuses two separate genes. This is supported by multiple available RNA datasets and our PFAM domain predictions, which revealed that the first half of the hypothetical protein encodes a 7TM_GPCR, while the second half encodes a p450 cytochrome. As it seems unlikely that a single protein would act as both, a GPCR and a cytochrome, we repeated our differential expression analysis after manually splitting PPA25752 into two separate genes. We found no evidence of differential expression for the resulting GPCR-gene, but could perfectly replicate the differential expression signature we originally found for PPA25752, for the cytochrome-encoding gene. FSOS = false sign or smaller, MMSE = minimum mean squared error, SRG = starvation-response gene