Check list
Before starting the pipeline, you can avoid most of the following errors by checking your input Sample sheet. With the newest version of the pipeline, this is check when initiation and creating the pipeline. Yet, you can do it yourself using this standalone application:
sequana_check_samplesheet --samplesheet SampleSheet.csv
Generally speaking, you should now the expected total number of reads of the sequeners (e.g. 400M reads for a NextSeq) and figure out whether the proportion of undetermined reads (not related to a sample) is reasonably low compare to the total number of reads.
Once the pipeline has finished, just open the summary.html file and check out the images
First, you should look at the summary plot. This shows the determined and undetermined reads. If the proportion of undetermined is large (e.g.>20%), this may indicate a wrong index sequence in your samplesheet. Here is a goo run:
This may also indicate a problem on a specific lane like in this example:
To get more insights into the most dominant undetermined indices, there is another plot available like the following one. With the same example as above, we see again that the lane 4 is causing trouble with an excess of undetermined in that lane only.
Finally, here is the same kind of plot for a NextSeq (4 lanes instead of 8) where everything went well and the rate of undetermined reads is very low.
