RNAhybrid outputting all 0 pvalues

[shedurkin]

rendered code RNAhybrid documentation RNAcalibrate documentation

I'm using RNAhybrid to predict miRNA-mRNA target interactions. One of the required inputs is parameter estimations for an extreme value distribution. RNAhybrid has built-in distibutions for 3'UTRs of humans, flies, and worms. However, since corals are evolutionarily distant from these options, I tried to use RNAcalibrate to estimate evd parameters from my data. When I ran RNAhybrid using these estimated evd parameters, all output hybridizations were assigned a p-value of 0. When I tried using the built-in 3utr_worm distribution though, I get a range of reasonable-looking p-values.

Example of all-0 pval output (pvalues bolded):

ntLink_8:665632-666632:999:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-22.8:0.000000:352: A C : GGAGAAAGUGCUG : UUUCUUUCACGAU :UG ACUAGUA ntLink_8:1175261-1176261:999:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-20.9:0.000000:772: C U : GUGCUGUGAUU : CACGAUACUAG :UGUUUCUUU UA ntLink_8:1175955-1176261:306:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-20.9:0.000000:78: C U : GUGCUGUGAUU : CACGAUACUAG :UGUUUCUUU UA

Example of output using built-in 3utr_worm distribution (pvalues bolded):

ntLink_8:32704327-32705327:999:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-27.8:0.036639:710:A A : ACAAGGAGAGUGCUGU : UGUUUCUUUCACGAUA : CUAGUA ntLink_3:84074-84461:387:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-26.7:0.014807:51: C U : GAGAGUGCUGUGGUU : CUUUCACGAUACUAG :UGUUU UA ntLink_6:2139830-2140830:999:Cluster_1826.mature::ntLink_6:4847465-4847486(-):22:-27.7:0.038633:949: G G : GGGGGGGGUGCUAUGAU : UUUCUUUCACGAUACUA :UG GUA