While the null distribution model may produce some inaccuracies for data with significantly non-Gaussian tails, we note that the 2-class model is unidentifiable without restrictions on the form of this distribution [37], and Gaussian distributions are widely used in error modeling.
We found that adding delay between the dose and the drug effect did not significantly improve the model fits, and we note that the time-scale for tumor shrinkage is on the order of months, whereas the terminal half-life of everolimus when dosed at 10 mg daily is around 30 h [ 9], so any delay is likely short compared with the time scales of interest in the study.
The acoustic startle response of the SCR mice was not impaired significantly, although we noted that the response with 110 dB sound tends to be reduced [Table S1; 110 dB, p = 0.0728, f(1,38) = 3.406; 120 dB, p = 0.1730, f(1,38) = 1.929].
We note that complete linkage significantly outperforms other linkage clustering algorithms: nearest neighbor linkage (single linkage) got a CH index of 14,042 and average neighbor linkage got 23,512.
Finally, we note that MRC strategies significantly increase the computational burden, since the calculation time scales linearly with the number of conformers.
We note that CPR2 is significantly more strongly violated in individual genes (exons in eukaryotes) than in non-coding regions (see, e.g., [24]).
First, we note that SiPhy identified significantly more sequences as constrained than both PhastCons and GERP.
We note that asparagine is significantly overrepresented in TRAD4, accounting for 29% of the amino acids.
Finally, we note that NGAL is significantly implicated in pregnancy, and was even named uterocalin in mice to reflect its high expression in the uterus [ 361, 537- 539].
We note that our algorithm is significantly different in structure from those of Karger and Ruhl, and perhaps substantially simpler.
