Argon belongs to the group of chemically inert noble gases, which display a remarkable spectrum of clinically useful biological properties. In an attempt to better understand noble gases, notably argon's mechanism of action, we mined a massive noble gas modelling database which lists all possible noble gas binding sites in the proteins from the Protein Data Bank. We developed a method of analysis to identify amongst all predicted noble gas binding sites, the potentially relevant ones within protein families which are likely to be modulated by Ar. Our method consists in determining within structurally aligned proteins, the conserved binding sites whose shape, localization, hydrophobicity and binding energies are to be further examined. This method was applied to the analysis of two protein families where crystallographic noble gas binding sites have been experimentally determined. Our findings indicate that amongst the most conserved binding sites, either the most hydrophobic one and/or the site which has the best binding energy correspond to the crystallographic noble gas binding sites with the best occupancies, therefore the best affinity for the gas. This method will allow us to predict relevant noble gas binding sites that have potential pharmacological interest and thus potential Ar targets that will be prioritized for further studies including in vitro validation.