short reads align to the assembled contig

[MonaLiu421]

Dear heng, thanks for the fantastic tool. I have some questions, can I use minimap2 to align pair-end 150bp short reads to assembled contig of these short reads, how is the performance? Or do you have any parameter suggestions? I remenber you said minimap2 doesn't work well on shorter than ~100bp reads. I have tried bwa-mem, however the larger the reference genome, the more memory is required (about 28*G).So i want reduce the consumption of memory by splitting the reference genome then align to every small ref, but one read may align to multiple positions, how do i select the best result of alignment? according the AS:i or XS:i or mapping quality？ Looking forward to your reply！

[lh3]

What is your input and command line options in use? I am surprised by the many lines having the same read name.

[MonaLiu421]

hi, heng. I'm glad to receive your reply, because I don't have so much memory to deal with large ref. I want to split large reference genomes like minimap2, divide them into small refs, and then combine all the alignment results. This is the merge BAM file. My command is as follows: 1. index bwa-mem2.avx2 index /data/input/assembly.fa the assembly.fa is the contigs set assembled by metagenomic short reads 2. alignment bwa-mem2.avx2 mem -T 30 /data/input/assembly.fa /data/input/AY2112.final_R1.fastq.gz /data/input/AY2112.final_R2.fastq.gz -R '@RG\tID:AY2112\tPL:illumina\tSM:AY2112' > /data/input/AY2112.sam after merge all alignments result, one read may align to multiple positions, i don't known how to select the best result of alignment? thanks, Looking forward to your reply!