The seqMINER standard analysis is separated in three steps from left to right as depicted in the image above. The green buttons are important, one need to click on these buttons for almost every analysis. You can do a multi-selection in a list by pressing the ctrl(separated selection) or shift(continuous selection) key and click the mouse key. The help message will appear on mouseover.
Step 1: Select your files
(1) select one reference coordinates file (i.e. peak file)
(2) select several aligned read files
The read files will not be loaded immediately, because this step takes the most of the analysis time. When you are sure to load the read files, select them and click on the "Load file(s)" button, they will be loaded simultaneously and appear in the list on the right.
Step 2: Extract data
When all datasets are loaded (the little green progress bar disappears), you can change the dataset order by clicking the up or down flash button. Loaded datasets can be deleted with the Delete button.
After all the green "Extract data" button will extract the density array from the datasets above according to the reference coordinates. Several parameters could be set in the option (see below).
Step 3: Clustering Visualisation
This step vary depending of the method used for analysis:
DENSITY ARRAY METHOD
Note: The clustering step can be skipped if you dont want to reorder your loci. Then you can visualize the heatmap using the Visualisation button.
A new item (distribution) will appear in the distribution list. Its name is composed by the peak file and the dataset name. You can rename it or export the raw data to a tsv file.
For the clustering, the main parameter to be set is the number of expected clusters (default 10) that you may want to modify according to the complexity of your dataset. The maximal number of runs(default 200) is reserved for advanced users.
Finally Kmeans clustering will be applied to the loci reorganisation and a heatmap view window will appear.
At that stage, you can reorder the clusters manually according to biological significance.
ENRICHMENT BASED METHOD:
The process is similar as above except that a single value is computed for each loci instead of an array of densities.
Then visualisation can be processed through dot plots.