7
$\begingroup$

I have to analyse two datasets consisting of time-series measurements of gene expression. One set are in vivo data from the expression profiles obtained between a few days before birth to several days post-partum. The second data set was obtained in a C2C12 cell line, where t = 0 was defined as the day in which 100% confluence was reached.

Is there an optimal way to align the two time series? Simply aligning the two series using the two t= 0 samples as starting point seems arbitrary.

$\endgroup$
  • $\begingroup$ Welcome to Biology! I attempted to clarify the question. Feel free to roll back if it doesn't reflect your question anymore. $\endgroup$ – AliceD May 28 '15 at 12:59
  • 2
    $\begingroup$ I would start asking myself: does it make sense for the two series to be aligned? In other words, do you have any evidence that the events in the in vitro experiment may correspond to events in vivo? It would help if you describe you experiment in a bit more detail, (if possible, this may be sensitive information). Which species, what tissue, what experimental conditions, what is your scientific question. $\endgroup$ – ddiez May 28 '15 at 14:28
  • $\begingroup$ Hi, yes, it is sensitive information... however thank you for the answer, I will try to find out if there is a best shift that globally maximises the correlation between pairs. If not everything will be traced back to up/down regulation... $\endgroup$ – Marco May 28 '15 at 14:45
  • 1
    $\begingroup$ This is a mesenchymal cell line that differentiates in vitro after reaching confluence (and serum deprivation). There are likely similar cells in the animals, but they might be a small percentage of the total number of cells. During differentiation in vitro you'd expect to see some genes going down and others going up over time. In the animals that signal would likely be swamped by tissue-specific genes from the dozens of tissues and hundreds of cell types. $\endgroup$ – mdperry May 28 '15 at 15:39
3
$\begingroup$

Direct comparison of the course would be difficult between tissues and cultured cells. In stead of time scale, you might want to use markers. For example, Kislinger et al. show the expression peak of SIX1 is 2 days after differentiation; SMAD3, 4 days. I am not a right person who can tell which markers are good, but I believe you could find more information by literature search.

http://www.mcponline.org/content/4/7/887.full

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.