A ranking system could help researchers choose well-characterized, reliable antibodies. Credit: Science Picture Co/Getty Images

Biomedical experts plan to create a scoring system that will help researchers choose reliable antibodies for their experiments. The only problems: figuring out how such a ranking would work — and getting manufacturers to adopt the standard.  

The idea comes from a workshop hosted this week in Asilomar, California, by the Washington-based Global Biological Standards Institute (GBSI), one of several groups concerned that poorly characterized antibodies are a major culprit behind the irreproducibility of biology experiments. Antibodies are large, Y-shaped proteins that are supposed to bind to specified biomolecules, helping researchers to track and identify them. But reagents that detect the wrong target — or that don’t detect the right one — have led to false findings, wasted resources and acrimonious controversies.

The concept of ranking antibodies is appealing, but doing so for even a fraction of those on the market would be an enormous task, says Roberto Polakiewicz, chief scientific officer of Cell Signaling Technology, an antibody manufacturer in Danvers, Massachusetts. “It’s still not clear how feasible that would be or how it would be implemented.”

The aim is to start by proposing definitions of how an antibody’s performance could be validated, says GBSI head Leonard Freedman. The reagents will need to have separate ranking systems for different kinds of experiments. For example, an antibody that accurately detects a protein in cells that have been broken open can be ineffective if used to detect the same protein in intact tissue.

Researchers would still need to validate even high-scoring antibodies for their own experiments, Freedman says. But a scoring system would give scientists confidence that an antibody will work as expected for its intended use.

Market demand

Where researchers would look up the scores and the data behind them is still unclear, as is how such data would be vetted. But detailing plans for the scoring system is more important than figuring out how to put it in place, says Joshua LaBaer, a proteomics researcher at Arizona State University in Tempe and a member of the workshop steering committee. “Just creating and defining a standard will create a bar that will improve the field,” he says.

If a few top vendors adopt a common scoring system, the practice will catch on, predicts Carl Ascoli, chief science officer at Rockland Immunochemicals, an antibody manufacturer in Limerick, Pennsylvania. “Others will follow suit because the market will demand it.” Scientific funders would also look to these standards, he says.

Anita Bandrowski, an information scientist at the University of California, San Diego, hopes that manufacturers will agree not just to score antibodies but also to provide lists of antibodies that were discontinued because of poor performance. These could be compiled into a database and used to alert authors, reviewers and editors about the need to be more sceptical of experimental results involving those antibodies.

Bandrowski is one of a group of researchers who, just before the meeting, published a paper1 proposing five broad strategies for validating antibodies. The workshop organizers aim to draw on these for proposals for each of the major kinds of experiments that rely on antibodies in about six months. The ensuing debate about how scores are calculated, and what details must be disclosed by antibody vendors, will be fierce, Ascoli says. “There’s going to be a firestorm.”