September 12th, 2008 – by: Adrian_Liston
This talk yesterday was so bad. It was billed as "mathematical modelling of in vivo dynamics of CD4+ and CD8+ Foxp3+ regulatory T cells". It promised so much, but delivered so little. They had some peptide that they wanted to see what it did to regulatory T cells so they injected it, a scrambled peptide or vehicle alone into two mice per group. They then bled the mice each day for sixteen days and did CD4/CD8/Foxp3 staining. The resulting graphs just zig-zagged up and down, with each of the groups being identical. Anyone else would have just concluded that their peptide did nothing, but no, these people did "advanced mathematical modelling" on the data. They assumed that the data was reliable and each up and down meant that the vehicle was changing the immune response.
The said that the fact that all three groups went up and down together made it ultra-reliable (rather than assuming that each day's staining was slightly different). They then noticed a mathematical trend - on those days where the proportion of CD4+ cells that were Foxp3+ went up, the proportion of CD4+ cells that were Foxp3- went down! Completely ignoring the basic mathematical truth that the two populations are also going to have inverse changes because they have to add up to 100% they then said that their model proved that the vehicle treatment caused spontaneous pulses of Foxp3+ T cell proliferation which in turn pulse down the Foxp3- proliferation. They also said that there was a blip of CD8+Foxp3+ T cells on day two, which proved that CD4+Foxp3+ T cells require priming by CD8+Foxp3+ T cells before they can function. I have never see such a load of tripe! Incidentally, Porto is the home of tripe since 1415 when the citizens of Porto showed their support for Henry the Navigator by supplying his voyage to Morocco with all of their beef, keeping only the cow stomach for themselves to make tripe from. Since then Porto's citizens have been known as triperios (tripe-eaters).