Criteria for identifying tandem and segmental duplicates

Question

[kozhaki.seq]

There are previous reports available on the estimated tandem and segmental duplicated genes (TDs and SDs) in rice (http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0063551). But I am not sure about the reliability of the data. Hence,I would like to analyze TDs and SD genes (not blocks) in rice.

Here, I am unable to make a criteria for TD and SD estimation (a condition which gives better reliable data). I noticed different groups used their-own criteria. For ex, the previous work uses "identity was > = 30% and the alignment covered > = 70% of the protein length" for TD. In other work, > 94% identity is used for SD (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3179966/)

1. Can anyone give suggestion before my analysis? I mean a cut off for TDs and SDs identification separately.
2. If anyone know a software suit which deals such duplicated genes estimation, it would be helpful. (FYI: I found some published suits, but the webpage is not working).

Thank you.

Answer 1

[Hranjeev]

The typical criteria that defines a segmental duplication is that the sequence is at least 1kb and 90% identical to the duplicate.