Evaluation of Compass as a comparative mapping tool for ESTs using horse radiation hybrid maps.

This research evaluates the effectiveness of the Comparative Mapping by Annotation and Sequence Similarity (Compass) strategy in accurately predicting the loci of equine expressed sequence tags (ESTs). It found that modifying the orthologue selection parameters could improve the prediction accuracy, albeit at the expense of reducing the total number of predictions.

Background and Objective

• This research aims to evaluate the performance of the Compass strategy in estimating the location of expressed sequence tags (ESTs) for species with comparative gene maps.
• The concept is to utilize human genome sequence information and known marker information from radiation hybrid (RH) maps for equine chromosomes to predict the locations of ESTs in horses.

Methodology

• The researchers used the human genome sequence information from Build 35 and marker information from the RH maps for equine chromosomes ECA 17 and X. They predicted 162 ESTs to locations on ECA17 and 328 ESTs to locations on ECAX by choosing the ‘top blast hit’.
• To evaluate the accuracy of the predictions made by Compass, the researchers experimentally determined the locations of 30 ESTs using RH mapping analysis.

Results and Analysis

• The results indicated that 53% (16 out of 30) of the predicted ESTs on ECA17 and ECAX were correctly mapped.
• These results suggested a requirement for identifying expressed orthologous sequences to enhance the precision of EST prediction.
• The researchers reassessed the locus predictions with three different modifications to the Compass strategy’s orthologue selection parameters.
• The accuracy improved significantly when opting for the ‘top gene hit’ (72% accuracy) or the ‘top expressed gene hit’ (86% accuracy).
• When the researchers utilized the default Compass parameters with the UniGene database, their predictions were 96% accurate. However, this high level of accuracy was achieved at the cost of a significant reduction in the total number of predictions.

Conclusion

• The research concluded that with optimized prediction parameters, the Compass strategy can be a useful tool for predicting in silico map locations for large sets of ESTs from unsequenced animal genomes.