Overestimation: erroneous assignment of species


#1

Hello,

I have a question. When I see my results in the inspect window it is not clear to me how Megan is assigned the reads. I thought that it only assigned a read to a specie level if it accomplish what it is stipulated in the LCA parameter window. But as you can see in the next image, for example for a contig that has two different species available (same score), it is assigned to one of them! I thought that since it has two different species with the same score, it must be placed it at genus level. Can you help with this. All my results are like this and they are wrong, since all species are overestimate .!

taxAssig|607x311


#2

Please re-paste the image, doesn’t appear to be there.


#3

Hi Daniel,

Never mind, I already saw my mistake. As I told you, I used the weighted algorithm in the LCA with 80% in the option “percentage to cover”; that why in some cases even do there are reads that have the same score to two different species, they are assigned to one of them (the bigger one).

Best regards,