Automated analysis of protein subcellular location in time series images

[biointelligence]

biointelligence                  28th June 2010

Automated analysis of protein subcellular location in time series images

Image analysis, machine learning and statistical modeling have become well established for the automatic recognition and comparison of the subcellular locations of proteins in microscope images. By using a comprehensive set of features describing static images, major subcellular patterns can be distinguished with near perfect accuracy. We now extend this work to time series images, which contain both spatial and temporal information. The goal is to use temporal features to improve recognition of protein patterns that are not fully distinguishable by their static features alone.

Results: We have adopted and designed five sets of features for capturing temporal behavior in 2D time series images, based on object tracking, temporal texture, normal flow, Fourier transforms and autoregression. Classification accuracy on an image collection for 12 fluorescently tagged proteins was increased when temporal features were used in addition to static features. Temporal texture, normal flow and Fourier transform features were most effective at increasing classification accuracy. We therefore extended these three feature sets to 3D time series images, but observed no significant improvement over results for 2D images. The methods for 2D and 3D temporal pattern analysis do not require segmentation of images into single cell regions, and are suitable for automated high-throughput microscopy applications.

Availability: Images, source code and results will be available upon publication athttp://murphylab.web.cmu.edu/software

Regards,

Team Biointelligence

Biointelligence

Getting started with biosciences

Website: www.biointelligence.in

Email: in...@biointelligence.in
Tel: +91-731-420-2793

 

P please consider the environment before printing this email. Thank you.2