Topic 4

2/29/2012 5:02:47 AM #

One issue with proteomics and its percieved lack of producing a substantive 'breakthrough' is that there is apparently no real will for the mass spectrometrist i.e. the person who generated and understands the data to take that data back into the biological field. The literature is crammed with manuscripts containing perfectly acceptable 'global quantitative proteomics' papers, perhaps supplemented with the odd western blot, ELISA or (worse) 'pathway/interaction' map from the latest current trendy commercial software package. But the results get no further. The average proteomics lab is staffed by mass spectrometrists/chemists who aim to publish a paper based on the outline above and move on to the next biological system. there needs to me more will in the field to treat the proteomics as 'Figure 1', and actually try to generate new avenues of research from the data. The arguement that "it's in the public domain so biologists can look at it if they want" is invalid, since many non-MS or non-omics people won't understand it and get 'snowblinded' by the weight of data. that's not a criticism - dealing with large datasets correctly is an acquired skill that few have.

One way to do this is to put proteomics into the hands of the biologist. Whether this is via more robust MS instrumentation (i.e machines which do not require substantive knowledge of chemistry and physics to optimise and troubleshoot), cheaper hardware (albeit with potentially less functionality, yet 'fit-for-a-specific-purpose'), or the expansion and associated reduction in cost of non-MS technologies such as antibody arrays.

Nowhere is the more keenly observed that in the field of clinical proteomics, where there are thousands of profiling "we tested 10 samples and these ten proteins are potential biomarkers" studies around, yet they vastly, vastly outnumber the studies which have then looked at these ten proteins in a larger cohort (by MS or other means). This may be publication bias, but the point remains. Taking the results of shotgun experiments and running with one is seen to be high risk for labs dependent on regular publications to help generate grant income. If the funding bodies could remove some of that risk (or the percieved penalites of inevitable failure when trying to go from discovery to clinic), or provide dedicated streams for this work, there may be more of a driver to try and take these data towards clinical benefit.

Anonymous | Reply