The result of FastANI was different with Nucmer

[YouxinZhao]

Hi I am using fastANI to analyze the average identity between chromosomes of two closely related species with conserved chromosomal macrocolinearity. Only one chromosome was used for each analysis. And I have observed differences in average identity among chromosomes. However, when I use Nucmer with the -mum option, filtering with delta-filter -q, delta-filter -1, and delta-filter -r, and then calculating average identity, the differences among chromosomes are not significant. And the correlation with fastANI results was low. For different chromosomes, the results from FastANI and Nucmer are inconsistent. For example, for chrname13, the FastANI result is 85.220963, which is the lowest among all chromosomes, but in the Nucmer results, it was the highest.

Subsequently, I calculated the correlation between Nucmer's average identity * (mapping length) / (chr length) and FastANI results, and the correlation coefficient is 0.98.

I would like to know whether I should rely on the results from Nucmer or FastANI for assessing average identity, as this might imply different conclusions.

Looking forward to your reply

shuyoushouxi@126.com

Best wishes Youxin Zhao

[YouxinZhao]

According to fastANi's diagram, fastaANi's calculation is similar to the result of delta-filter-q