The experiment table shows a list of currently defined experiments.
Each experiment has a name, which is initially Exp nnn, where nnn indicates a unique number. The experiment can be renamed by clicking in the name field.
The view also shows the number of samples that were selected from the plate view for this experiment, as well as the number of probes analyzed.
When an experiment is selected, its probes, samples and variables are shown in tabbed set of tables underneath the experiment table. All the charts for the experiment are drawn and put into a series of tabbed panes to the right. The bargraph pane is brought to the front.
- The "norm" column gives the type of normalization chosen for this experiment.
- The "background" column gives the type of background subtraction chosen for this experiment, either "water" or "probes" or "custom" if both negative wells and blank probes have been designated.
- The "MDD" column gives the type of minimum detectable dose calculated for this experiment, either "probes" or "water". If blank probes are being subtracted, the MDD is considered "custom".
The following actions are driven by the experiment table and its right-click popup menu.
- The names of the experiments can be modified by clicking and typing. Names must be unique.
- Experiments can be merged by selecting one or more and using the popup. There are three kinds of merge, and to understand them, it is helpful to remember that an experiment has a set of input datapoints and a set of output datapoints, with the output being calculated via the data processing
- An intersection merge. The resulting experiment will have the union of the samples and the intersection of the probes of all the input experiments. Background, normalizer and standard curve information will be preserved in the resulting experiment. The experiment output will be result of applying the data processing pipeline to the combined input.
- A union merge. The resulting experiment will have the union of the samples and the union of the probes of all the input experiments. Missing entries will be filled in with empty data points. Background, normalizer and standard curve information will be preserved in the resulting experiment. The experiment output will be result of applying the data processing pipeline to the combined input.
- A multi-plate merge. The resulting experiment will have the union of the samples and the intersection of the probes of all the input experiments. The output data points from the input experiments are used, and negative, blank, normalizer and standard curve samples are removed so that no further processing is carried out on pre-processed input. This option is useful for combining standardized data from multiple plates, where it is desired to apply the standard curves of each plate to that plate's samples. (A regular merge would pool all the standard curve information and build a single standard curve for all samples.)
- Experiments can be duplicated. This is often useful prior to deleting probes (or samples) in an experiment, to compare results with and without the extra probes.
- The replicates in an experiment can be merged, creating a new experiment. Each data point in the resulting experiment has a statistic formed from by collecting the means of each replicate data point; for instance if three replicates with 42 particles, 24 particles and 39 particles were combined, the replicate sample would have 3 "particles" representing the mean of each replicate.
- Experiments can be deleted
The Negative Button and Normalize Button apply to the currently selected experiment.
The Export Button exports all experiments.
Help Front Page