How to feed HISAT2 results into expression tools?

Prepare HISAT2 Aligned Reads

• HISAT2 generates SAM format output by default. First, convert the SAM file to BAM format using SAMtools for efficient processing.

 

• Sort the BAM file by coordinates using SAMtools, as many expression tools require sorted BAM files.

 

• Index the sorted BAM file using SAMtools to facilitate quick access to the data during downstream analysis.

Select an Expression Quantification Tool

• Common expression quantification tools include StringTie, featureCounts, and HTSeq. Choose one that best fits your needs in terms of features and compatibility with downstream analyses.

 

• Ensure your chosen tool is properly installed in your environment. Follow specific installation guides for each software to avoid compatibility issues.

Set Up Annotation Files

• Download the reference annotation file (GTF or GFF format) for your organism. This file contains genomic feature information necessary for counting reads.

 

• Ensure the annotation file is compatible with your BAM file in terms of reference names and versions.

Run Expression Quantification Tool

• For StringTie: Run StringTie with the sorted BAM file and the annotation GTF file to estimate expression levels. Use the `-e` option to enable more accurate expression estimates by limiting read assignment to specified features.

 

• For featureCounts: Run featureCounts with the sorted BAM file and the annotation file. Ensure you specify appropriate options for strand-specific counting if your data require it.

 

• For HTSeq: Run HTSeq-count with the sorted BAM file and the annotation file. Specify options for stranded data if necessary, and choose the appropriate mode such as intersection-strict or union.

Output and Format Results

• For StringTie: The output will be a GTF file containing expression levels. You can convert this file into a tabular format using additional StringTie utilities if needed.

 

• For featureCounts: The output is a tab-delimited text file listing counts for each feature. You may further process this file to normalize counts using R packages like DESeq2 or edgeR.

 

• For HTSeq: The output is a counts file, which you can feed into statistical analysis packages for differential expression analysis.

Post-Processing and Quality Checks

• Check the alignment rates and feature counting statistics to ensure data quality. This helps identify issues such as low mapping rates or unexpected biases.

 

• Perform additional normalization and batch correction if necessary, using tools and methods suitable for your data and study design.

Integrate into Downstream Analyses

• Use the expression data in downstream analyses such as differential expression analysis, clustering, or pathway analysis. Ensure your data is appropriately normalized and corrected for systematic biases before further analysis.

 

• Document all data processing steps for reproducibility, and consider sharing your pipeline and scripts within your research community.