A high-profile 2013 study that concluded that different kinds of happiness are associated with dramatically different patterns of gene activity is fatally flawed, according to an analysis published on Monday that tore into its target with language rarely seen in science journals.