Should boostDM exceptions be handled as errors?

[koszulordie]

We might consider whether common boostDM exceptions are exited as errors or not.

For example, it is questionable whether an exception such as not finding a high enough quality model for a given gene-ttype pair should raise an error or, alternatively, just exit(0) the program with an appropriate logging.

Even though nextflow is told to ignore errors, every error in the pipeline produces an annoying printout. In the case of processes such as PredictSaturation where many exceptions are expected, the excessive amount of printouts can be disturbing, even if the run is completely fine.