CrossFilt: A Cross-species Filtering Tool that Eliminates Alignment Bias in Comparative Genomics Studies.

Area of Science:

• Genomics
• Bioinformatics
• Comparative genomics

Background:

• Comparative functional genomics studies face challenges due to biased read mapping across species.
• Inter-species differences in genome structure, sequence composition, and annotation quality contribute to mapping biases.

Purpose of the Study:

• To develop and validate a filtering strategy to improve the accuracy of cross-species gene expression analysis.
• To ensure read count quantification is based on directly comparable genomic features.

Main Methods:

• Developed CrossFilt, a filtering strategy that retains only sequencing reads mapping reciprocally between genomes.
• Utilized real and simulated RNA-sequencing data from primates for evaluation.

Main Results:

• CrossFilt demonstrated superior performance compared to five alternative approaches.
• The strategy resulted in more accurate inference of gene expression differences in primates.
• Showcased the impact of preprocessing strategies on cross-species functional genomics data analysis.

Conclusions:

• CrossFilt provides a robust method for accurate cross-species gene expression analysis.
• Reciprocal read mapping is crucial for reliable comparative genomics.
• Preprocessing strategies significantly influence the outcomes of functional genomics studies.