# How to read E-value annotation on NCBI BLAST?

i am a bit confused how to read the current E value annotation on NCBI BLAST results.

I looked into the matter but could not find an factual answer .

What i understood so far: E -values are the expected value of the number of random querys of the same length as the input that would result in a even or better S-score on the same sized database. Since E-values are estimates, not probabilities , they can be lower than 0.

However, if NCBI BLAST retuns an e-value of say: "4e -19" do they mean:

• 4*10^⁻19 (this would not result in negative values )
• or do they mean 4 E-19 as in : " four querys score better in S with - 19 difference"
• It's a number in regular scientific format, so it's the former. 4e-19 = 4 * 10^-19 Feb 9 at 18:38
• thanks for the reassurance! .. man, sometimes i stumble upon one source that uses a weird notation and i start to get quite confused.. The source of my confusion was Victor Flores comment on researchgate.net/post/… Feb 9 at 19:16
• After rereading i could have seen earlier that he is apparently wrong. Feb 9 at 19:18
• He never says anything that resembles your second interpretation. How did you end up with that interpretation? Feb 10 at 22:16
• You are totally right, i confused "beyond" with "below" (not a native english speaker) . This should never have resulted in such a stupid confusion, unfortunately the feedback at university is a little bit sparse due to the pandemic circumstances , so you had to bare with my undergrad incertitude. I honestly thank you for clearing it out nonetheless. On the upside i learned my lesson, took a step back and studied Altschuls formulation. Feb 13 at 12:17