How to deal with substring in contigs in metagenomic research?

Question

[hophuquy0944]

In my metagenomic study, I encountered issues constructing a feature table from contigs generated during assembly. I am concerned about potential substring relationships between contigs, as these could introduce bias in feature counting. Could you provide insights on handling such cases and advice on constructing an accurate feature table from contigs? Thank you for your assistance.

[dthorbur]

Have you noticed there is a bias in results, or this is just a hunch? If the substrings in question are repetitive elements, you can mask them using something like repeatmasker.