ABySS Bloom filter false positive rate

Question

[PeiwenLi]

Hi! I am using abyss-pe with the Bloom filter mode to perform genome assembly. The false positive rate (FPR) was reported twice in the log file as shown below: ... Loaded 478640000 reads into Bloom filter Loaded 478650000 reads into Bloom filter Loaded 478658235 reads from /home/peiwen/WGS_2018/1181_S2_L004_R2_001.fastq.gz into Bloom filter Bloom filter FPR: 3.94% Counting Bloom filter stats: #counters = 190887435392 #size (B) = 190887435392 threshold = 2 popcount = 13530549696 FPR = 0.00252% ...

I wonder which FPR I should keep track of? And how are the two different?

Thanks! Peiwen

Answer 1

[onestop_data]

Looks a like a question for https://github.com/bcgsc/abyss/issues