After sending out samples of synthetic peptides to over 100 laboratories, the Association of Biomolecular Resource Facilites is gearing up to conclude the Proteomics Research Group’s fourth study, which focuses on identifying de novo sequences.
“Sometimes protein sequences are not in the database, if you’re working with certain species whose genomes are not very-well sequenced, or if there are errors or post-translational modifications,” said Thomas Neubert, an assistant professor at New York University School of Medicine who chairs the ABRF Proteomics Research Group. “It’s important sometimes to determine the sequence of a protein without having the sequence in a database.”
Since its inception in the 1980’s, ABRF has conducted numerous studies, most of which are concluded within one year, to address specific biological problems. Study samples are sent out to anonymous laboratories, which investigate the samples on a voluntary basis. Results are then sent back to leaders of the research group who compile and present findings at an annual ABRF meeting, and later publish the results in a peer-reviewed journal.
This year, the Proteomics Research Group chose to focus on peptides whose sequence is not found in any public database to make people aware of the problem and to find out what are the best solutions to the problem, said Neubert. The study will conclude on Dec. 13. Results will be presented at ABRF’s annual meeting to be held Feb. 5-8 in Savannah, Ga., and will most likely published in the Journal of Biomolecular Techniques.
Neubert said it will be interesting to see how different types of mass spectrometers fare with de novo sequencing. Also, some laboratories may use Edmund degradation to do the sequencing.
“A couple of years ago we used to have to do this quite a lot because the sequences were pretty patchy and proteins were often not in databases,” said Neubert. “We’d have to do de novo sequencing, and then do BLAST searches against sequenced genome to see what came up.”
William Lane, the director of Harvard University’s Proteomics Analysis Facility and ABRF’s Executive Board liaison to the Proteomics Research Group, said that the current study will evaluate instrumentation, algorithms, and expertise in deducing a sequence without having a database reference.
Doing de novo sequencing via mass spectrometry is not a straightforward task, said Lane. “Mass spectrometry has exquisite sensitivity and power, but it’s still a database-dependent technology, by and large,” he said. “De novo sequencing by mass spectrometry typically requires further interpretive skill, and the Proteomics Research Group study will evaluate this.”
In addition to evaluating instrumentation, the study will also let researchers know which software, if any, has been particularly helpful in analyzing unknown sequences.
“Some people won’t use any software and they’ll get good results; others may use software and not get good results — we’ll just have to see what the results are,” said Neubert.
This year is the fourth year since the inception of ABRF’s Proteomics Research Group. Before the group was founded in 2001, ABRF had a Peptide Synthesis Group, a Protein Identification Group, a Protein Sequencing Group, a Peptide Standard Project Group, and a Mass Spectrometry Group.
Last year, the ABRF Proteomics Research Group chose to study the differentiation of protein isoforms by sending out samples with a mixture of three closely related, intact samples. Eight out of 42 labs correctly identified all three isoforms, while seven out of 42 did not identify any proteins.
Researchers concluded that separating proteins before analysis did not influence the success rate of the analyses in this simple protein mixture, and that use of enzymes other than trypsin was not useful for most analyses. In addition, they concluded that relying on protein-identification software was not sufficient to make accurate discrimination between closely related isoforms. Instead, careful inspection of raw MS data was usually required.
In 2002, the Proteomics Research Group chose to study posttranslational modification by sending out samples of phosphorylated proteins and asking participants to analyze the sample and identify the sites of phosphorylation.
Only one laboratory correctly identified the two phosphopeptides in the sample correctly, as well as their phosphoryalation sites, and two other labs correctly determined both phosphorylation sites. Researchers concluded that despite advances made in instrumentation and techniques, locating sites of phosphorylation in cold samples is still a significant challenge.
“These studies are usually quite informative and useful,” said Neubert. “It’s not a contest or a competition. We really think that we can learn by getting the results from everyone.”
Neubert said that this year’s study gives a good opportunity for researchers who have never done de novo sequencing before to give it a shot.
“I don’t think anyone is going to invent a new method just to solve this problem, but I think this study introduces labs to new techniques they wouldn’t have tried,” said Neubert. “Everyone in the field can learn from others in the field.”