Fit lines do not follow the sample points in plotGenes

[dosshra]

Hello
I have plotted the trends 4 most significant DE genes in my analysis using plotGenes function (see attached file). Three plots show peaks of either case, control, or combined that do not seem to follow the counts of the samples.
trend_plot_3_1stPage.pdf

Thank you
Hanan

[davidsebfischer]

Hi Hanan, this looks ok to me, the overfitted maxima lie do not cover any observations and do not effect the rest of the fit too much. Note also that the impulse model will not result in perfect interpolation like splines do because it introduces a pretty strong prior on the shape of the signal, this also makes it more robust though.