Dear Listers, I've applied basic shape statistics to a natural oval slab of tissue in which developmental patterning is known to occur along both major and minor axes. I've been thinking of the intersection of early gene expression and shape of structures within the tissue slab. I'm going to attempt a text diagram below of the direction I'm thinking about, to see if my problem resonates with any others on the list. The diagram below will look OK as along as monospace fonts are used.
Let's say the tissue slab is thus: +--------A--------+ | | | | | | | | | | | | | | | M--------+--------L | | | | | | | | | | | | | | | +--------P--------+ and up down is anteroposterior and left right is mediolateral. I have a some anatomically defined fields in the tissue slab as thus: +--------A--------+ | | | -------- | | \ / | | \ / | | \ / | M ++++ \/ L | + + | | + + | | ++++ /\ | | / \ | | / \ | +--------P--------+ I know from pilot work that these fields differ in size and shape between genetic lines, and thus likely have a genetic causal component. I know from developmental work that this system is similar to body plan development, in that transcription factors act as morphogens during early development, perhaps in gradients along the major and minor axes. My approach to this problem will be forward genetic, in the sense that I want reliable measures of the phenotype (shape) with with to correlate extant genetic polymorphism in experimental crosses of genetic lines to isolate QTL affecting shape. On that front I feel confident, and have a grant under review. Thus, I might have the following hypothetical biological result: +--------A--------+ | | | -------- | | \ / | | \ / | | \ / | M ++++ \/ L | + + | | + + | | ++++ /\ | | / \ | | / \ | +--------P--------+ = genotype BB @ X, N = 40 genetic lines, 5 animals per line +--------A--------+ | ********** | | \ / | | \ / | | \ / | | **** \ / | M + + \/ * L | + + /| | | + + / | | | ++++ / | | | / | | | / | | +--------P--------+ = genotype DD @ X, N = 40 genetic lines, 5 animals per line I'm less clear about expanding in a direction to meld expression genetics (aka "genetical genomics") with shape differences, but I believe this is a very important direction. With regard to gene expression phenotypes I will be likely limited (economics and dissection skills) to samples of tissue that divide the slab into four quadrants from each animal from each genetic line: +--------A--------+ | | | | | | | 1 | 2 | | | | | | | M--------+--------L | | | | | | | 3 | 4 | | | | | | | +--------P--------+ It is then relatively easy to find QTL affecting anteroposterior and mediolateral differences in gene expression. It is also then possible to look for cis-QTL in this data set and the same QTL in the shape data set above for overlap. But I'd also like to directly associate gene expression differences by quadrants with shape differences. I have not yet figured out the appropriate Monte Carlo analysis to assess power or the most efficient statistic association methods to employ, but I'm working on that. Is anyone else thinking on similar problems?? -Dave -- David C. Airey, Ph.D. Research Assistant Professor Department of Pharmacology School of Medicine Vanderbilt University 8148-A Medical Research Building 3 465 21st Avenue South Nashville, TN 37232-8548 TEL (615) 936-1510 FAX (615) 936-3747 EMAIL [EMAIL PROTECTED] URL http://people.vanderbilt.edu/~david.c.airey/vita/ URL http://www.vanderbilt.edu/pharmacology -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
