Skip to content

benchmarking.md

Latest commit

 

History

History
181 lines (145 loc) · 4.42 KB

benchmarking.md

File metadata and controls

181 lines (145 loc) · 4.42 KB

Benchmarking SequelTools

We used Maize line NC358 data to benchmark all tools of SequelTools. Specifically, eight SMRTcells of PacBio Reads were used (both scraps and subreads).

m54138_180610_050652.scraps.bam		 m54138_180610_050652.subreads.bam
m54138_180610_152029.scraps.bam		 m54138_180610_152029.subreads.bam
m54138_180619_053104.scraps.bam		 m54138_180619_053104.subreads.bam
m54138_180620_123320.scraps.bam		 m54138_180620_123320.subreads.bam
m54138_180620_225343.scraps.bam		 m54138_180620_225343.subreads.bam
m54138_180621_091000.scraps.bam		 m54138_180621_091000.subreads.bam
m54138_180627_120814.scraps.bam		 m54138_180627_120814.subreads.bam
m54138_180627_222429.scraps.bam		 m54138_180627_222429.subreads.bam

1. Benchmarking QC tool

A. With subreads only

A file of file names (fofn) was generated for subreads (one per line) labelled subreads.txt. A run script (bash) was used to test QC tool on various number of processors (runSequelT-QC-sub.sh).

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -u subreads.txt \
    -t Q \
    -p a \
    -g a \
    -n ${cpu} \
    -o runSequelT-QC-sub_${cpu}_8-SMRTcells

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-QC-sub.sh $i
done

B. With scraps only

A file of file names (fofn) was generated for subreads (one per line) labelled scraps.txt. A run script (bash) was used to test QC tool on various number of processors (runSequelT-QC-scr.sh).

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -u subreads.txt \
    -t Q \
    -p a \
    -g a \
    -n ${cpu} \
    -o runSequelT-QC-scr_${cpu}_8-SMRTcells

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-QC-scr.sh $i
done

2. Benchmarking Filter tool

Filter tool requires both subreads and scraps, so subreads.txt and scraps.txt, previously created fofn was used to test filtering tool based on (a) Adapters, (b) Passes, and (c) Size. The benchmarking was done similarly as explained above (8 SMRTcells and 4-16 CPUs).

A. Adapters

The run script was set up as follows (runSequelT-F_adp.sh)

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -c scraps.txt \
    -u subreads.txt \
    -t F \
    -N \
    -n ${cpu} \
    -o runSequelT-F_adp_${cpu}_8-SMRTcells
rm runSequelT-F_adp_${cpu}_8-SMRTcells/*.sam

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-F_adp.sh $i
done

B. Passes

The run script was set up as follows (runSequelT-F_pas.sh)

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -c scraps.txt \
    -u subreads.txt \
    -t F \
    -P \
    -n ${cpu} \
    -o runSequelT-F_pas_${cpu}_8-SMRTcells
rm runSequelT-F_pas_${cpu}_8-SMRTcells/*.sam

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-F_pas.sh $i
done

C. Length

Fixed length (1000bp) was used for length filtering. The run script was set up as follows (runSequelT-F_len.sh)

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -c scraps.txt \
    -u subreads.txt \
    -t F \
    -C \
    -Z 1000
    -n ${cpu} \
    -o runSequelT-F_len_${cpu}_8-SMRTcells
rm runSequelT-F_len_${cpu}_8-SMRTcells/*.sam

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-F_len.sh $i
done

3. Benchmarking Sub-Sampling tool

For Sub sampling the subreads, we used length based sub-sampling method (select longest subread per CLR). Since it only needs subreads, we used the subreads.txt fofn as input. The benchmarking was done similarly as explained above (8 SMRTcells and 4-16 CPUs), using the runSequelT-S_len.sh script.

#!/bin/bash
PATH=$PATH:/ptmp/GIF/arnstrm/sequeltools/SequelTools/Scripts
ml r-devtools samtools python
cpu=$1
time ./SequelTools.sh \
    -u subreads.txt \
    -t S \
    -T l \
    -n ${cpu} \
    -o runSequelT-S_len_${cpu}_8-SMRTcells
rm runSequelT-S_len_${cpu}_8-SMRTcells/*.sam

For running it on 4 to 16 CPUs:

for i in $(seq 4 16); do
  runSequelT-S_len.sh $i
done