Complexity curve

1. Looking into complexity curve. Could it be that for PE reads we actually need to add the -P parameter in preseq? This was not detected in the pipeline automatically.

samtools sort 6-10A_S13_L001_R1_001_val_1_bismark_bt2_pe.bam \
         -m 8589934592 \
         -@ 1 \
         -o 6-10A_S13_L001_R1_001_val_1_bismark_bt2_pe.sorted.bam
     preseq lc_extrap -v -B 6-10A_S13_L001_R1_001_val_1_bismark_bt2_pe.sorted.bam -o 6-10A_S13_L001_R1_001_val_1_bismark_bt2_pe.ccurve.txt 

2. Also, for very small amounts of the reads this program is not working, but I guess that is my problem with reads of total number 119,573. Error: max count before zero is less than min required cound (4), sample not sufficiently deep or duplicates removed.

original command:

BAM_INPUT
TOTAL READS     = 119573
DISTINCT READS  = 119456
DISTINCT COUNTS = 3
MAX COUNT       = 3
COUNTS OF 1     = 119342
MAX TERMS       = 2
OBSERVED COUNTS (4)
1       119342
2       111
3       3

ERROR:  max count before zero is les than min required count (4), sample not sufficiently deep or duplicates removed

and if I implement PE:

PAIRED_END_BAM_INPUT
paired = 119572
unpaired = 0
MERGED PAIRED END READS = 119572
MATES PROCESSED = 239144
TOTAL READS     = 119572
DISTINCT READS  = 119485
DISTINCT COUNTS = 2
MAX COUNT       = 2
COUNTS OF 1     = 119398
MAX TERMS       = 2
OBSERVED COUNTS (3)
1       119398
2       87

ERROR:  max count before zero is les than min required count (4), sample not sufficiently deep or duplicates removed

Start by reviewing the pipeline entry point that assembles the preseq complexity-curve command, especially handling of paired-end BAM reads and the possible -P parameter. Reproduce the shown commands with the small read counts and determine whether the expected behavior is correct or needs a documented pipeline change.