Identifying protein function--a call for community action.

Abstract

During the last few years, we have seen enormous strides in our abilities to sequence genomes, and the information that has poured out of these sequences is quite astonishing. With more than 150 complete genome sequences now available and many laboratories rushing into microarray analysis, proteomic initiatives, and even systems biology, it seems an appropriate time to consider not just the opportunities those sequences present, but also their shortcomings. By far the most serious problem is the quality and degree of completeness of the annotation of those genomes. Most troublesome are the large numbers of open reading frames that have been identified by computer programs, but remain labeled as a “conserved hypothetical protein” when they occur in more than one genome or simply a “hypothetical protein” when they appear unique to the genome in question. Between them, these two categories of annotated open reading frames often represent more than half of the potential protein-coding regions of a genome. These annotations highlight just one portion of our ignorance about the information content of genomes and our lack of fundamental knowledge about the function of so many of the building blocks of cells. Unless we rectify this situation, it is likely to undermine many of the other “-omic” efforts currently underway. Here I advocate a rather straightforward approach to address this problem—focused initially on the bacterial genomes. In contrast to the numerous proposals for big science initiatives to understand the fundamental workings of biological organisms, I propose a small science, relatively low-tech approach that could have a dramatic pay off. A relatively small investment could yield a massive amount of information that would greatly enhance our current efforts to use genomic approaches to study life.