ASCB Newsletter Dec 2013 - page 17

17
DECEMBER 2013
ASCB
NEWSLETTER
The ability to reproduce the results of published
research is fundamental to the scientific
enterprise. I recently encountered an example
of how frustrating it can be for researchers
who discover that even results published in
prestigious journals may not
be reproducible. I was giving
a workshop about effective
manuscript writing to a group
of postdoctoral fellows and
their mentors. We started
by critiquing abstracts from
published papers that I thought
were particularly well written—
informative and intriguing. Later,
when we discussed Methods
and Results, I mentioned that
Nature Publishing had introduced a new
“standards checklist”
1
and was asking authors
to document where in the manuscript they had
addressed each checklist item. The checklist
posed questions such as whether the cell lines
and antibodies used in the study had been
authenticated, how many times a study was
repeated in the laboratory, whether the assays
were done in a blinded fashion, and what
statistical tests were used and
why
those tests
were selected. The idea behind the checklist is to
make the research methods reported in
Nature
journals more transparent,
with the aim of making
the work more robust and
reproducible.
After my talk, a senior
scientist thanked me for
mentioning the checklist. And
then, pointing to one of the
abstracts we had discussed
earlier, he said, “we tried to
reproduce the work in that
paper,” which had been
published in a highly regarded
journal. “We tried, but we
could not replicate the findings in that paper!”
Irreproducibility in biomedical research has
gotten lots of attention recently, even making
the front page of
The Economist
.
2
More and
more, researchers are voicing concerns like the
one I heard at the manuscript workshop: that
many published papers—often in prominent
journals—contain results that cannot be
independently replicated. Consequently,
time, money, and, sometimes, precious
biospecimens or rare reagents are being
wasted.
Setting Standards
My own awareness of these problems
shot up several notches when I attended
the 2013 Biotechnology & Standards
Conference, co-sponsored by the
Massachusetts Institute of Technology’s
Center for Biomedical Innovation
and the Global Biological Standards Institute
(GBSI), a new organization affiliated with
ATCC.
To help frame the standards issues, including
the problems of irreproducibilty, it is necessary
to clarify what is meant by standards. The
CEO of GBSI, Leonard Freedman, did so,
distinguishing between material standards
and procedural (written) standards. Material
standards refer to well-characterized, purified
biological reagents such as reference cultures
of cells; any specific RNA, DNA, or proteins
derived from those cells; or any
chemically synthesized versions of, or
variations on, such macromolecules.
Procedural, written standards refer
to documentation of specifications,
procedures, or processes. As Freedman
and other conferees mentioned, the
development and implementation of
material and procedural standards may
take different routes and be overseen by
different entities and therefore should
be dealt with separately.
An important example from the
Standards conference was the problem in the
biomarker sphere. Life scientists and biomedical
researchers have been making major efforts to
identify predictive and prognostic biomarkers
True and/or False: Biomedical
Publishing Standards Need
Upscaling
“We tried, but
we could not
replicate the
findings in
that paper!”
Career Adv i ce for
Women and Men
WOMEN
in Cell Biology
Beth Schachter
1...,7,8,9,10,11,12,13,14,15,16 18,19,20,21,22,23,24,25,26,27,...48
Powered by FlippingBook