search

henryxushi / BayesDenovo

1 stars 0 forks source link

"Beginning kmer hash ... Killed" #1

Open RNiloP opened 2 years ago

RNiloP commented 2 years ago

Dear Henry

many thanks for this useful tool! I tried the test file and got the files "both.fa", "bayesdenovo.sam", "bayesdenovo.fa", "assembly.fa" and "Assemble.log", as expected. However, when running a larger pair of fastq files, the program ended with the following message:

"Error, cmd: /PATH/BayesDenovo-v1/src/Assemble --reads both.fa -k 25 --bamfile /PATH/BayesDenovo-v1//Ccelo/output/bayesdenovo.sam --fafile /PATH/BayesDenovo-v1//Ccelo/output/bayesdenovo.fa --pair_end --double_stranded_mode --debug 2>Assemble.log died with ret 35072 !"

PATH = /media/david/90aba6e6-45be-4827-9400-ae7db2b6022c/ (I did not write it to simplify the output).

When checking the "Assemble.log" file, you can read:

"Begin loading reads ... 19397930 reads have been loaded ! Beginning kmer hash ... Killed"

I am not sure what could have happened. I am attaching the bash "run_BayesDenovo_Ccelo.sh" as a txt file.

run_BayesDenovo_Ccelo.txt

By the way, I am running a desktop with Ubuntu 20.04.2 LTS and Python 3.6.13 .

Many thanks!

Ricardo

PD1. A not directly related question, is it possible to merge 3 different fastq files before running the regular BayesDenovo pipeline? I have a right_1, right2 and right 3 fastq files, and their corresponding left files.

henryxushi commented 2 years ago

Hi Ricardo, thank you for trying our package. I think the problem is related to the memory of your machine. Can I know the memory specification of your machine? De novo assembly tool is very memory consuming. It would be better to run on computing clusters with large memory configuration (>20 GB). For your question, yes, you can merge the fastq files before running the pipeline. Thanks.

RNiloP commented 2 years ago

Dear Henry

Many thanks for the prompt reply. I am using a DDR4 16GB RAM, CPU Core i7-5930K PC.

henryxushi commented 2 years ago

Dear Henry

Many thanks for the prompt reply. I am using a DDR4 16GB RAM, CPU Core i7-5930K PC.

I think the problem will be the RAM. If it is possible, please run on a machine with at least 30 GB RAM.

RNiloP commented 2 years ago

Ok, I will try to upgrade to a PC with 32 GB, and let you know how it worked. Best!

Ricardo