Why my `Percent duplicates` in a shallow sequencing is too high?

Question

[631079064]

I made a 10x snATAC-seq library and do a shallow sequencing of 5G. Typically, the Percent duplicates is less than 20%. However, my library had 78% duplicated. Did the shallow sequencing do something with this ?

Here are my web_summary.html, and most of warnings can be attributed to the shallow sequencing.

Answer 1

[LauferVA]

Yes, probably, albeit possibly indirectly.

It seems that a few duplicates are dominating the profile of your output.

This probably relates to two things based on the output above; first, the number of reads in cells/non-cells, which ultimately would relate to depth if depth is low enough.

Second, depth wouldn't solve the problem of duplicated reads, but it likely relates the the distribution of output reads you got by ''cutting the experiment short," so to speak, in the sense that the few reads you had wouldn't have accounted for so much of the product if other products had also amplified well, which ultimately would have occurred if the depth had been high enough to allow it.

Please verify this last comment: I think you might be able to circumvent this if you

• titrated the primer concentration better. this would have the effect of limiting the max number of duplicates you could obtain per cycle at an earlier cycle number than was the case currently.
• lessened the number of targets (library size)

Anyway its hard to know without thumbing through the data, but yes, probably.

[631079064]

Thank you for your help !

[631079064]

It means I should change the primer concentration and use beads to control library size to reduce the duplicate rate. However, I follow close to the 10x snATAC protocol .....

[LauferVA]

Do you have a good relationship with 10x? Do you have a representative? Honestly, I would ask them, then post the results here.

I'd be interested to know what they say.

VL

[LauferVA]

It means I should change the primer concentration and use beads to control library size to reduce the duplicate rate. However, I follow close to the 10x snATAC protocol .....

did you follow the snATAC seq protocol for primer concentration RELATIVE TO THAT READ DEPTH though? What I am saying is that the proportion of those might be out of sync.

However, there are a lot of other possibilities here. The results could relate to specimen preparation or cell type or some such. What is your preparation like?

Are you running it at a much lower read depth than they recommend? If so, why?

[631079064]

I follow the 10x protocol with an engineer from the local 10x company. But the library had very low peak signal at 400 & 600 bp . So I decided to do a shallow sequencing according to the engineer's advice.