"By comparing our findings with the Yooseph clusters, we also discovered hundreds of new gene families that hadn’t even been seen before," says Zhai and adds that by adding the diverse GOS data to known profiles, "we were able to make them more sensitive and diverse, and so increase their power to categorize novel sequences."
Diversity of microbial kinases
In a separate study, Manning, Zhai, and first author Natarajan Kannan, Ph.D., a postdoctoral researcher in the lab of HHMI investigator and UCSD professor Susan S. Taylor, Ph.D., traded the breadth of the ocean survey for the depth of a single protein domain. They zoomed in on kinases, extremely well studied enzymes, which control every aspect of eukaryotic cell biology and are important cancer drug targets. They control the activity of proteins and small molecules by attaching tiny phosphate groups to them. By contrast, much less has been known about their bacterial counterparts.
Again and again, the researchers combed the GOS data for bacterial kinases, each time rebuilding their domain profiles by including the new members found in the previous round. All in all, they dug up 45,000 protein kinase sequences that fell into 20 distinct families, of which the eukaryotic protein kinases are just one. The additional 19 families spanned a huge range and included several that had never been described before.
"Prokaryotic protein-like kinases were considered to be some sort of niche players, but actually they are more prevalent and widespread than histidine kinases," explains Manning. Bacteria were thought to rely mostly on histidine kinases, which are structurally different from protein kinases, for all their signaling