Fastq Stat

Fastq Stat is a high-performance tool designed to process and analyze FASTQ files, which are commonly used in bioinformatics to store nucleotide sequences and their corresponding quality scores.

Prerequisites

Before building and running the Fastq Stat program, ensure you have the following prerequisites installed:

• Intel oneAPI DPC++/C++ Compiler: Required for compiling the program. You can download it from the Intel oneAPI Toolkit.
• libdeflate library: Used for efficient decompression of gzip-compressed FASTQ files. You can clone and build it from the libdeflate GitHub repository.
• large size memory: It loads two fastq.gz files into memory and decompresses them. So, it loads two fastq.gz and two fastq files into memory at the same time.
• avx512bw: cpu that supports avx512bw

Architecture

The program is written in C++ and leverages the Intel oneAPI DPC++/C++ Compiler for optimal performance. It uses the libdeflate library for efficient decompression of gzip-compressed FASTQ files. The program is designed to take advantage of modern CPU features, such as AVX-512, to accelerate computation.

Build

To build the Fastq Stat program, use the following command:

/opt/intel/oneapi/compiler/latest/bin/icpx -o fastq_stat fastq_stat.cpp ./libdeflate/build/libdeflate.a -lpthread -std=c++20 -O3 -mavx512bw

This command compiles the fastq_stat.cpp source file, links it with the libdeflate library, and optimizes the code for AVX-512 capable processors.

Execute

To execute the Fastq Stat program, use the following command:

./fastq_stat {input1.fastq.gz} {input2.fastq.gz} {thread_count}

• {input1.fastq.gz}: Path to the first input FASTQ file.
• {input2.fastq.gz}: Path to the second input FASTQ file.
• {thread_count}: Number of threads to use for processing.

Example

Here is an example of how to run the program:

./fastq_stat ../EPG24-AAMY_1.fastq.gz ../EPG24-AAMY_2.fastq.gz 10

This command processes the EPG24-AAMY_1.fastq.gz and EPG24-AAMY_2.fastq.gz files using 10 threads.