What to do with data that has low read depth

Question

[stat.1405]

Read Depth.

In the data which has this structure

pos      ind1-1     ind1-2     ind2-1    ind2-2
1          20         0         15         100
2          0          1          2          50

There is 20,000 position or call variants because every position means SNP. There are 10 individual every individual has 2 replicates. When I looked at the DP, I found there are many ZEROs. according to many papers they call it gap. because there is no read in this position, what I understand from the data above every replicates has its won read depth.

What can I do with low read depth. maybe it will increase the false positive??

Answer 1

[Devon Ryan]

Normally you would filter by some minimal read depth, since any calls resulting from low-depth are unreliable.

Answer 2

[stat.1405]

I found the answer, in VCF files if you have no read (0) , the genotype will be ./. (missing)