mritchielab / FLAMES

A framework for performing single-cell and bulk read full-length analysis of mutations and splicing.
https://mritchielab.github.io/FLAMES/
GNU General Public License v3.0
17 stars 9 forks source link

Error in locating minimap2 #20

Closed sh0rt2l0ng closed 3 months ago

sh0rt2l0ng commented 8 months ago

I tried to feed blaze2 fastq into flames. But got below error.

Error in if (file_test("-f", file.path(minimap2_dir, "k8")) && file_test("-f", : missing value where TRUE/FALSE needed In addition: Warning messages: 1: In base::system2(command = "command", args = c("-v", "minimap2")) : error in running command 2: In base::system2(command = "command", args = c("-v", "minimap2")) : error in running command 3: In base::system2(command = "command", args = c("-v", "samtools")) : error in running command

I tried to evaluate k8 and paftools.js manually. Both are "TRUE". Not sure what caused this.

file_test("-f", file.path(minimap2_dir, "k8")) [1] TRUE file_test("-f", file.path(minimap2_dir, "paftools.js")) [1] TRUE

Could you please show some light on this? Thanks!

youyupei commented 8 months ago

Hi @sh0rt2l0ng,

Thanks for being interested in FLAMES, to help us better understand the issue, would you be able to share the R code and config file you used?

Thanks

sh0rt2l0ng commented 8 months ago

Hi @youyupei ,

Thanks for the reply! Here is the R code and config file generated with the code.

config_file.json flames_test.txt

youyupei commented 8 months ago

Hi @sh0rt2l0ng,

It seems the minimap2_dir was not passed into the sc_long_pipeline function. It might be helpful if you add minimap2_dir=minimap2_dir when calling sc_long_pipeline function. i.e.,

sce <- sc_long_pipeline(annotation = annot, fastq = fastq, genome_fa = genome_fa, minimap2_dir=minimap2_dir,
                          outdir = outdir, config_file = config_file, expect_cell_number = 5000)

Also some suggestions: in your config file, you could

  1. use more threads by modifying the number (depends on how many cores you have) in "threads": [1] to shorten the run time.
  2. set "bambu_isoform_identification": [true], for potentially more accurate isoform identification
sh0rt2l0ng commented 8 months ago

Hi @youyupei ,

Thanks for your advice. I will definitely try this and let you know the result later. BTW, may I know what's the minimum hardware requirement of FLAMES? I am going to use it for a 150GB fastq.gz from ONT-10X protocol. What's the peak memory usage of each thread in FLAMES? Or which step in FLAMES consumes the most memory in run time.

youyupei commented 3 months ago

Hi @sh0rt2l0ng , not sure you still need this information but just in case, from my experience ~150G of fastq file, the peak memory can go to ~60-80G. The steps consume most memory would normally be the mapping step.