Stein Gives Bioinformatics Ten Years to Live
Subject:   longevity of bioinformatics
Date:   2004-09-24 05:03:17
From:   wjbug
Response to: so nice you posted twice?

Hello All,

First, let me pass on thanks to O'Reilly for sponsoring this conference and specifically to Mr. Steinberg for passing on his view of Dr. Lincoln Steins presentation.

I think there's a fairly easy way to decide what tone is appropriate for posting comments to a venue your scientific colleagues are likely to peruse.

I) Express yourself in a manner commensurate with an editorial comment in a scientific journal - e.g., would you want to be quoted in a primary scientific publication such as 'Nature' and 'Cell' or a scientific news journal such as 'Science News'.

II) Would you use the same tone were you to have attended the conference and made a similar comment directly to the speaker.

We must assume the statement you are commenting on was put out there by Dr. Stein himself, rather than being an editorial comment made by the author of this article, as it appears in quotes:

"No, they're studying life. Biologists like Ernest Mayer can sit in his office and look at other people's data and develop theories of selection. When people ask me, I say I'm a biologist."

Had the author gone to the web in order to find the correct spelling of Dr. Mayr's name, I expect you'd have been much less likely to adopt such a tone. There are many references to this revered investigator available on-line, including the library of the eminent Museum of Comparative Zoology at Harvard which is named in honor of Ernst Mayr (

Having made such a preface, I hope I'm now able to express my own criticism cogently.

I agree with the general content of this anonymous posters comment. I assume my criticism ought to be addressed to Dr. Stein, as the comment is given as a quote.

First, it is a bit incendiary to express this observation in such a manner. People familiar with Dr. Mayr's work are likely to have strong feelings about whether they agree or not with his deeply considered theories. Whether they emphatically agree or equally emphatically disagree with Dr. Mayr's theories/observations, the notion a seminal contributer to the field would "...sit in his office and look at other people's data..." is likely not to strike anyone in his field as truly reflecting the nature of Dr. Mayr's contribution even over his post-septuagenarian years.

There is a significant difference between performing novel experimental field and/or benchtop research and working to cast the body of this research in a coherent theoretical framework. Both means of contributing to a scientific field are absolutely essential to the progress of scientific knowledge. I believe this is essentially the point Dr. Stein was seeking to express in his extemporaneous reply to a question after his talk.

My bigger concern, however, is with what I assume based on this report is Dr. Stein's general presentment bioinformatics will be absorbed into the various biological fields to which it is relevant. This appears to be based on the argument "bioinformatics" is a tool analogous to a microscopy or PCR. My difficulty with this argument is these latter "tools" are based on specific physical principles which - broadly defined are:

  • microscopy: diffraction, transmission, reflection and/or emission of EMF radiation of various frequencies of to produce description of a material based on its physico-chemical properties
  • PCR: temperature-cycling-induced annealing of polynucleotides leading to the amplification of specific individual nucleotide sequences either for the purpose of detecting or producing said sequences.

Though both of these techniques have extremely varied means of application, those general physico-chemical principles are in effect in all these situations.

Bioinformatics can be broadly defined as:

  • bioinformatics: computer-aided organization, manipulation and/or analysis of biological scientific data

    Not only does this definition extend well-beyond the specific application of computational tools to bio-molecular sequences & structures - the application subject to a decade-long, massive proliferation starting in the late 1980s - it is also so general a category as to be nearly unlimited in its applicable problem domain. Whereas technical innovations - and thus novel experimental observations - using microscopy or PCR need to provide a new means to address the aforementioned physico-chemical principles, the properties being exploited in bioinformatics data manipulation appear to be constantly expanding. Witness the proliferation in the past several years of principles culled from the confluence of computational linguistics, artificial intelligence and database research leading to the wide-spread application of knowledge-frameworks, ontologies and controlled-vocabularies as a means to identify and perform correlative analysis on the semantic qualities of biological data.

    I believe this leads to two results:

    • the spread of bioinformatic applications to an ever-increasing collection of computational disciplines;
    • the extended presence of the various "-ics" - genomics, transcriptomics, proteomics, pharmacogenomics, physiomics, behaviomics (, ecolonomics, evolutiomics (sic) - over the coming century. I believe the existence of these technical disciplines will not only continue through the century, they are likely to uniquely inform & rejuvenate various other biological disciplines such as systems biology, multigenic analysis and complex ecological modeling.

    Just my $0.02.

    Thanks again for covering the conference.

    Bill Bug
    Laboratory of Bioimaging & Anatomical Informatics
    Dept. of Neurobiology & Anatomy
    Drexel University College of Medicine

  • 1 to 1 of 1
    1 to 1 of 1