Paperonium

Protein evolution · eLife · 2024

“Higher than random chance”: 12 of 17. Chance does that 1 time in 14, too often.

1 in 100every time1 in 20the usual barabout1 in 41The paper: 13 or more, passesabout1 in 14Its own setup: 12 or more, fails
1 in 100every time1 in 20the usual barabout1 in 41The paper:13 or more, passesabout1 in 14Its own setup:12 or more, fails

How often chance alone gives a result this strong if touching does nothing

The paper says

Do touching parts of a protein machine evolve in step (speeding up and slowing down together) more than other parts? In 12 of 17 machines they did. The paper’s test says chance alone would give that only 1 time in 41 (p = 0.0245).

Its own data show

By the paper’s own setup, if touching does nothing each machine is a coin flip: 12 of 17 is like 12 heads in 17 tosses. Chance alone gives 12 or more heads 1 time in 14 (p = 0.072), missing the usual 1-in-20 bar. The paper’s 1 in 41 counts only 13 or more.

This test asks how often chance alone gives a result at least as strong as the one seen, so 12 itself counts. Leaving out exactly 12 heads (1 time in 21) turns 1 in 14 into 1 in 41. Corrected, only one of the paper’s three checks that touching matters passes. Its main finding, no general link, stands.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Evolutionary rate covariation is a reliable predictor of co-functional interactions but not necessarily physical interactions · eLife (2024) · doi:10.7554/eLife.93333