It is a whole lot much easier to find a solution when the problem is well understood. Available in silico tools afford scientists today this unique approach of first understanding the biochemistry of the disease before venturing into drug discovery.
As any calculation, in silico methods has a probability to fail in predictions made by. I think it is a good starting approach, but it should be reinforced with experimental data. Metabolic diseases as any others could have multiple processes involved and computational tools are not afforded yet to include all of this in consideration. So limitations are there, but it does not imply that this approaches be wrong, we just need to improve them.
I think there's a limitation in everything. If you are planning an experiment, you will always have limitations. I think the point is that you have to consider this lack of certainty as part of your results.
Christian Alonso Becerra You have a point. Beatrice Pecoraro Taking into consideration the number of drugs targets and leads available and the number of drug approvals, we should start considering quantifying the uncertainties... may be some mathematical formulae could be useful.