no ANI results + AAI Fraction of proteins shared (%) ="(estimated)"May 23, 2019 at 1:40pm
Hi, I recently run three of my assemblies through MIGA, and for one of them I got weird results regarding ANI and AAI. First, for one of the assemblies there is not ANI table to view. Does is mean that MIGA found no good hits for my genome? Second, when looking at AAI values for the same record, some of the records have to following string in the 'Fraction of proteins shared (%)' column: (estimate). What does it mean, how it should be interpreted. Best, Mikolaj
May 23, 2019 at 2:59pm
- MiGA only calculates the ANI between pairs with AAI ≥ 85%. If your dataset doesn't report an ANI table, it means that all of the AAI values were below that threshold.
- In order to speed-up the searches, MiGA first attempts an estimation of AAI using only a set of genes conserved in Archaea and Bacteria. When these genes reach 90% identity (corresponding to roughly 70% AAI), the estimate becomes unreliable, and the complete AAI is performed. This estimation is robust, but the fraction of proteins shared cannot be estimated directly from these, therefore the "(estimate)" instead. If your AAI table ONLY contains "(estimate)" values in that column, it's because all of the matches were below 70% AAI.
Thanks a lot for these questions! We'll be working on incorporating this and some extra information directly in the documentation. Please don't hesitate to ask any follow-ups or any other questions.