which statistics to apply?

Purpose of my study is to understand better how the experimental lines I am using compare to the positive and negative controls under untreated or treated conditions and if the differences are meaningful.

I have treated and untreated plates. The setup of an untreated and treated plate are the same and looks like this: Positive_Control|Negative_Control|Experimental_Line_1. I have 4 experimental lines, so I have 4 untreated plates and 4 treated plates in this kind of setup. I do not want to compare experimental lines to each other, just to the controls on the same plate. Thus, I have 3 independent groups in a plate.

My sample size is small, less than 15, on each plate. I repeated the experiment three different times so I have about 30 sample points per line now. I want the samples samples to reflect a whole population. I may think of pooling the repeats together, but I want to see how the individual trials look first.

I have calculated mean, SD, and SE for the positive, negative, and experimental lines and have compared the lines in a preliminary analysis by looking at if the means of the experimental lines were bigger or smaller than the controls. My hypothesis is that on untreated plates, the means are bigger than the negative control and on treated plates the means are smaller than negative control. This has been the case in the real experiments.

I am not sure what statistics method to choose, whether it be a student t-test or ANOVA and which specific kind under either (unpaired, parametric, etc).

What advanced statistics is right for this? Any suggestions will help.
OK, I have made some progress over the last post. I decided to do ANOVA one-way for groups more than 2, and unpaired t-tests on groups that have only 2.

My next question is that I have 2 controls: positive and a negative control plated along with my 2 experimental lines. I know Dunnet's test in ANOVA will compare a control to other groups but it only has 1 control. Are there any tests that can use 2 controls?