Bio-Question Driven Bioimage Informatics Workflow Design Using Bio-TDS

Biological images (bioimages) acquisition using high-throughput screening technology has led to a large increase of dataset amount in bioscience research labs. However, management infrastructures (collect, store, transfer, share, etc.) and analysis remains the bottleneck in most research project. The analysis challenge include providing the community with relevant analytic tools at each step of the bioimage informatics workflow development based on user questions of interest. Bio-TDS (Bioscience query Tools Discovery System) provide a free text ontology annotation approach to discover the most relevant Bioinformatics tools from user question. This allow a researcher to accelerate the image informatics workflow design by asking successive stepwise questions that lead to the analysis goal. For example, you can ask question such as “how can I compare two or more cell confocal microscopy images?” From this first step question, you can decide if you will like to compare your images by sub-region (using ttorg) or compare the cell network structure (using cellset). This first choice will lead to a new step in the workflow with a new question such as “how accurate is the 3D cell network structure projection from image?” Bio-TDS suggest to use im2net to build the 3D projection of your image. We tested Bio-TDS and compare with existing tools repositories and Bio-TDS outperform any state of the art retrieval system both on the precision and the recall.

Reference

1. Gnimpieba,E.Z., VanDiermen,M.S., Gustafson,S.M., Conn,B. and Lushbough,C.M. (2016) Bio-TDS: bioscience query tool discovery system. Nucleic Acids Res., 10.1093/nar/gkw940.