Predicting Errors in Gene Expression Data from Confocal Images
Author Information
Author(s): Myasnikova Ekaterina, Surkova Svetlana, Stein Grigory, Pisarev Andrei, Samsonova Maria
Primary Institution: St.Petersburg State Polytechnical University
Hypothesis
Can a regression system accurately predict errors in gene expression data obtained from confocal images based on microscope parameters?
Conclusion
The developed regression system can accurately predict and correct errors in gene expression data from confocal images, even for strongly clipped images.
Supporting Evidence
- The regression system was trained on images from three different microscopes.
- The method allows for correction of errors in data obtained from strongly clipped images.
- The software tool CorrectPattern is freely available for use.
- The study demonstrated high prediction accuracy in error estimation.
Takeaway
This study created a tool that helps scientists fix mistakes in pictures of tiny organisms, making it easier to understand how genes work.
Methodology
The study developed a regression system trained on images from three different microscopes, using PMT gain and offset as independent variables to predict error sizes.
Limitations
The method requires a representative learning sample from the same confocal system and scanning experiment for accurate predictions.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website