extreme increase in kraken2 runtime

[ma-yaqin]

Hello,

So I am running some data sampled from respiratory tract to explore the microbiome. So far, all samples run smoothly with runtime ranging from 10-45 minutes using my current AWS instance (32 threads, 72 GB RAM, 2 TB volume disk). however, I had this one particular sample (let's call it sample A) that runs for about 2 days now which is odd. to put perspective in the matter, circumstance is as follow:

1. number of reads in sample A is actually similar to that of other samples
2. I used memory mapping since putting database in RAM is not possible
3. before the analysis of sample A, I was running out of volume disk so I had to increase it to 2.5 TB (gp2, SSD)
4. database used is standard database (2022)

here is the code I used in case it is needed. kraken2 --db /data/blastdb_temp/krakendb --threads 30 --memory-mapping --paired sample_A_1.fastq sample_A_2.fastq --report sample_A_paired_kraken2_report.txt > sample_A_paired_kraken2_classification.txt

However this code works well on other samples. So does anyone have idea on this? any help is appreciated. thank you in advance

[dhjcp3]

中国兄弟吗？你是怎么建数据库的啊 我的下载老是失败 能交流一下不

[Solenyalyl]

中国兄弟吗？你是怎么建数据库的啊 我的下载老是失败 能交流一下不

中国兄弟在这儿，数据库可以单独下载，获得了manifest.txt之后，在里面加上ncbi链接的前缀和wget -c，就可以下载了