Curve pattern identification and categorization

3 posts / 0 new
Last post
bokcmho
bokcmho's picture
Curve pattern identification and categorization

I'm working with HTS generating time course of Ca signals. The changes of Ca level is complex, goes up and down during the first few minutes. I've also found that the ways the Ca level changes represent a specific signature of the particular response, or a characteristics of particular type of receptors (e.g. very different patterns from GPCRs to ion channels, or to RTKs).

Here come to my questions. I want to fit a curve to the particular pattern, or some way to identify each pattern of the time course of Ca signal changes. Then use this categorization to screen for any similar pattern triggered by unknown compounds / receptors. What I want to achieve is a novel drug screening strategy based on the shape and duration of Ca signals generated by the ligands on a particular system.

Anyone could suggest me for a solution? I'm using VBA to try to achieve this in Excel, but it's not a trivial job...

qiang wang
qiang wang's picture
it seems to me that you have

it seems to me that you have a multi binding of Ca2+ to your ligand, in this case, I feel a Gauss peak fitting will work it out.

bgood
bgood's picture
You may find that you have

You may find that you have many more tools at your disposal if you shift away from development with VBA/excel and towards development with something like Java (e.g. the [url=www.cs.waikato.ac.nz/ml/weka/ ]weka[/url] algorithms) or R.