Epidemiology: Slow Metabolism of Alcohol Linked to Lower Heart Risk Genetics: Public, Private Drafts of Genome Found Comparable Neurology: Early Decision: How Embryonic Stem Cells Become Fine-tuned Brains Neuroscience: New Center Will Bring Basic Neuroscience to the Bedside Public Health: New Vaccines Could Balance Global Burden of Disease Digital Library Update Gene Initiates Joint Formation Radiation Limits Narrowing of Arteries After Stent Growth Factor Seen to Reverse Loss of Muscle from Aging, Disease T Cell Response to HIV Proteins May Make Them Vaccine Candidates Barger Awards Honor Outstanding Mentors, Silen Award Recognizes Lifetime of Mentoring Candidates Sought for HMS Dean for Continuing Education HSPH Holds Poster, Exhibit Day Women's Health Conference Seeks to Increase Participation of Minority Women in Clinical Trials HST Events Highlight Biomedical Technology, Student Research Clinical Scholars Take Master's in Patient-oriented Research Front Page

GENETICS Public, Private Drafts of Genome Found Comparable But How Will They Develop as Research Resources? In the race to sequence the human genome, the conflicts between the two contestants have often overshadowed the prize itself. During the week of Feb. 12, Celera Genomics published its draft version in Science while the Human Genome Project's own publicly funded version appeared in Nature. An important consequence of the dual efforts is the existence of separate sequences and their implications for the scientific community. The proof is in the pudding: A team led by George Church (front) found that after the much-hyped conflict between public and private, the two resulting versions of the human genome proved to be pretty equal. With Church are (l to r) Jay Shendure, Adnan Derti, and John Aach. Photo by Graham Ramsay George Church, HMS professor of genetics; John Aach, HMS lecturer on genetics; Jay Shendure, an MD–PhD student; and colleagues had the opportunity to perform one of the first comparisons of the two draft sequences, detailed in Nature. They found that the rivalry of the public and private enterprises has spurred and improved a potentially sluggish public project. "The managers in the public domain are highly motivated to undermine the utility and cost-effectiveness of any private resource," said Church. This motivation has produced a public draft that is comparable to the private one. "In terms of quality of assembly and quality of data, they're surprisingly similar," he said. "You can see how quickly a lead can be eroded in this field." Productive Rivalry The sequences may be similar in quality and size, but they are assembled differently. The public project generated a series of overlapping clones that were sequenced separately and reconstructed based on the overlaps and other information. Celera used a whole-genome shotgun sequencing approach to bypass the need for overlapping clones. But just as Celera's efforts galvanized work in the public domain, the company owes much of its progress to the freely available public data, which it used liberally to supplement its own data. There are years of studies to be done to analyze, compare, and refine the draft sequences, but Church and his colleagues began with a few relevant evaluations. "The various tests we used to compare these genomes were biased toward things we thought biologists would want to be doing as complete genomes come out," he said. To measure the completeness and continuity of each draft, the group looked at gaps in the assemblies, which are indicated in the sequences by strings of N's. At first, it seems as if the sequence of the Genome Project contains fewer unidentified bases than the Celera sequence, but this is because Celera was able to more accurately predict the size of the gaps and includes larger stretches of unidentified sequence. Another measure of a sequence's continuity is how likely it is that a particular stretch of DNA will be found on a continuous piece, or "contig." Both sequences are split into contigs of various sizes, far short of the ideal of having a single contig for every chromosome. The team looked at 10 large mRNAs, which are encoded by a series of exons that can be spread out over a long distance. "If the mRNA is on more than one contig, that probably means the assembly isn't complete, so you can use that as an assay for how complete the assembly is," said Church. The sequences yielded similar results, but both had some examples of incomplete mRNAs or ones that spanned multiple contigs. The team then looked at oligonucleotides, unique short sequences that can be used as probes and primers for specific stretches of DNA or mRNA. "To set up for the rest of the biological annotation of the human genome, it helps to know which sequences are truly unique," said Church. The team looked at all possible unique combinations of 15 nucleotides, or 15-mers, and found that the two drafts had comparable numbers, but there was a small percentage that were found in one draft and not the other. In these cases, having two drafts can serve to check one against the other, with the stretches found in both drafts making likelier candidates for study. Beyond gene hunting, one of the challenges for scientists is to interpret the 99 percent of the genome that does not code for proteins. These stretches contain many DNA binding sites that serve as docking stations for transcription factors and may be located far afield of the genes they regulate. "The challenge is that a lot of these factors act in concert with one another and have fairly degenerate and small sites compared to protein motifs." The team searched for binding sites in upstream sequences that were likely to contain the motifs. Again, the two sequences were comparable, but there still were discrepancies between them. The journals both include several papers that begin the long task of applying and interpreting information from the drafts. These sequences now form the most detailed map of the genome and must be linked to other kinds of map that sketch out the borders and landmarks that will make the data navigable. Integrating Maps Bradley Quade, HMS assistant professor of pathology at Brigham and Women's Hospital, was part of the BAC Resource Consortium, which published a study linking the public sequence to chromosomal landmarks. These landmarks, large chunks of DNA carried in bacterial artificial chromosomes and each containing sequence tags that had been inserted like mileposts into genomic sequence, were mapped onto every chromosome. The product, an integrated map of the human genome, will be an important tool for studying chromosomal rearrangements associated with tumors and congenital abnormalities. Quade said that having the map makes looking for genes a much faster process. "In situations where you have a piece of positional information, it's hard to translate it to where you are on a sequence map," he said. "It's like miles versus kilometers." It is this sort of annotation that Church believes will make the public database more useful in the long term. He believes that Celera will still attract researchers who are willing to pay the $15,000 subscription fee to have a head start in their field, even if there is only a small chance they will find what they need in the private database. But the public database will become a repository for the interpretation of data that will make the genome more useful. "Some may afford the private database, but when they go to annotate it, they'll go to the public one. That's where all of the energy of the researchers is going to go." —Courtney Humphries