Albina Asadulina: ImageJ in the workflow for generating, evaluating and visualizing 3D gene expression atlas

Abstract

Gene expression atlases were developed to overcome the limitations of traditional single- or multi-color labelling methods and light microscopy which are often restricted in the number of channels (and thus genes). Gene expression atlases show spatial distribution and co-expression of genes, revealing potentially interacting genes.

We present an automatic procedure for generating gene expression atlas that is based on the published image registration protocol. The new protocol subsequently applies rigid orientation using ImageJ and non-rigid transformations using the ITK toolkit. We demonstrate the method on the example of Platynereis dumerilii. The advantage of Platynereis is that its larvae at the same stage exhibit consistent morphological features that can be used for the registration. Some of these features can be determined using ImageJ therefore we developed a custom algorithm for rigid orientation of the Platynereis images. This allowed us to speed up the registration process.

We generated four averaged anatomical templates, one based on the acetylated-tubulin signal and one based on DAPI stained nuclei, for both two and three day old larvae. For that we iteratively register and average 40 individual stacks for each template.

We then registered the expression patterns of cell-type specific genes to the generated templates. In order to evaluate the accuracy of the registration of gene expression patterns we analyzed the average absolute deviation of cell-center positions using ImageJ. We found that both templates allowed cellular-resolution gene expression registration in whole-body scans, with the DAPI-based template often performing significantly better than the acetylated-tubulin-based one. We also determined that 3-5 scans are sufficient for the registration of an individual gene.

To visualize a gene expression atlas, we developed a Channel Merger plugin for ImageJ. It allows merging multiple greyscale stacks into one image.

Keywords

Image registration, 3D gene expression atlas, Platynereis

Short CV

I am a PhD student in the Jékely Research Group (Neurobiology of Marine Zooplankton) at the Max Planck Institute for Developmental Biology, Tuebingen, Germany, funded by the Max Planck International PhD Program. I hold a Diploma in Software Engineering (Kyrgyz-Russian Slavic University, Kyrgyzstan, 2008) and MCs in Life Science Informatics (Rheinische Friedrich-Wilhelms-Universität Bonn, Germany, 2010). My research interest is analyzing biological images. Currently, I am working on the marine worm Platynereis dumerelii, for which I am generating a gene expression atlas using image registration.

Administrative data

Presenting author: Albina Asadulina
Organisation: Max Planck Institute for Developmental Biology

co-authors: Albina Asadulina, Aurora Panzera, Csaba Veraszto, Gáspár Jékely