distinguish the paired-end reads

[zhangzhiyangcs]

Hi,
I have some questions:

1. I have many samples. if I merge two fq in one fq for each samples, it will cost many time and disks. Do you have any suggestion to improve the command "sed 's/ /_2 /' FASTQ2|cat FASTQ1 - > merge.fastq" that I can imput 1.fq.gz and 2.fq.gz directly.
2. Why use tophat2 instead of STAR or hisat2?
3. I have annoted 2400 cscRNAs in one sample. Is there an scripts to filter the false positive cscRNAs?
   Best wish

[wyt14]

we test result: tophat2 has the relativelay high accuracy.

[wyt14]

The filter script will be updated soon~