First, I want to make clear a strong point. It seems likely that the people working with Murthy did not know their data was fabricated. Several of them I know personally, and they are strong enough personalities to stand up to such things.
Still, they have been negligent because the WHAT_CHECK software dis exist already in those days; it was published in Nature in 1996. WHAT_CHECK produces a series of strong warnings that something is seriously rotten with many of the Murthy files.
I have so far kept this very simple check secret. But, had any of the Murthy post-docs Emailed me (like many people do who find anomalies in WHAT_CHECK reports), I would in most cases have advised them strongly not to publish.
For example, 1bef, 1dfg, 2ou1, 1l6l, 2a01, and 2qid have all residue distribution characteristics of membrane proteins. It would be nice, from a homology modellers point of view, to know which software Murthy used to find his homology modelling templates, and how he did the modelling (which software). The fact that all these proteins look like membrane proteins suggests that his modelling skills are equally poor as his crystallographic skills.