Skip to main content

Table 3 Runtime of our implementation

From: TMRS: an algorithm for computing the time to the most recent substitution event from a multiple alignment column

Train, gradient (1 iteration)100 K columns4.6 min
Train, total (300 iterations)100 K columns23 h
\(t_{\text {MRS}}\), \(\sigma\), q1 column\(7.3\times 10^{-4}\) s
\(t_{\text {MRS}}\), \(\sigma\), q1 G columns204 h
  1. We show runtimes of our implementation. We used 100 species vertebrate multiple alignments for the measurements. For training data, we used a sampled alignment with 100 K columns from 4d sites. As for the computation of \(t_{\text {MRS}}\), \(\sigma\), and q, we used the sampled alignments from 3\('\)UTR sequences which have 2,034,681 total alignment columns, and scaled the runtime for each Datasize