One person’s take on how to write a paper in the

One person’s take on
how to write a paper in the
life sciences
John Boothroyd, PhD
Professor of Microbiology and Immunology
With some edits and comments by John Pringle
Professor of Genetics
WHY publish?
• Make your work public for the greater
good (otherwise, why did you bother?)
• Establish your reputation as a scientist
• Establish your reputation as a
responsible grantee who delivers value
for the funds expended
• It FEELS so good!
What really matters!
• Accuracy of the data presented!! (Without
this, nothing else matters much!)
• Rigor and clarity with which the data are
described and interpreted
• Intellectual honesty in relating the work to the
published literature
• “A reputation takes years to build and
seconds to lose…”
Why write well?
• If there has obviously been care expended in
making the writing clear and mistake-free,
there is a natural tendency to suppose that
the research has been done with equal care.
And vice versa!
• Good writing makes your story easier to
understand, which makes potential readers
(1) more likely to read it, (2) more likely to
understand it, and (3) less likely to be
annoyed with you. (All of which are good!)
Sad (or maybe not) but True!
• A mediocre thinker and/or
experimentalist who writes very well is
likely to have a successful independent
• A brilliant thinker and/or experimentalist
who writes poorly is unlikely to have a
successful independent career (although
s/he may do well as a student and
First steps
• Begin the process way before the work is
complete to identify “logic holes” AND so
that the job is less daunting later
• Decide on a tentative title to help
crystallize what the paper is really about
• Decide on target journal and look
carefully at its format (print a paper)
• Sketch out figures and tables (rough)
• Decide on authors (with advisor and
• ALL those who:
– contributed data that are presented in the paper
– or served an important, on-going intellectual role in
the design and interpretation of the experiments
• Not those who simply funded the work or
offered a few helpful suggestions
• Not those who just “need another publication”
• Order is the responsibility of the senior author
• Joint first and last authors possible
• For Heaven’s sake, don’t be petty.
(Remember, “A reputation takes years to build
and seconds to lose…”)
• Generally, succinct declaration of conclusions
reached rather than methods used (unless
basically a methods paper)
• But … must be FOREVER an UNDENIABLY
TRUE statement based on DATA SHOWN (and
how many conclusions reach that point??)
• OK to temper with “Evidence for…”, “Putative”,
“Apparent”, “Role of…” if not iron-clad
• Noun descriptor use sequence avoidance
recommended (i.e., avoid using sequential
nouns as adjectives)
Which would you prefer?
• “Mapping effector genes by microarray
analysis of human foreskin fibroblast cells
infected with F1 progeny of a cross between
two strains of the obligate intracellular
Apicomplexan parasite Toxoplasma gondii.”
• “Toxoplasma co-opts host gene expression by
injection of a polymorphic kinase homologue.”
Good Titles!
• Campbell,…, Boothroyd (1984) Nature
“Apparent discontinuous transcription of
Trypanosoma brucei variant surface-antigen
• Sutton ad Boothroyd (1986) Cell “Evidence for
trans-splicing in Trypanosomes”
• Short and sweet
• 1-2 sentences to set the stage [describe system;
articulate what was unknown when began]
• 1-2 sentences to say what was done
• 2-3 sentences of what was observed
• 1-2 sentences of conclusions reached
• No references (but depends on journal format)
• No unexplained abbreviations
• No unsubstantiated conclusions
• Carefully choose key words for search discovery
Example Abstract
“The majority of known Toxoplasma gondii isolates from
Europe and North America belong to three clonal lines that
differ dramatically in their virulence, depending on the host
[=intro]. To identify the responsible genes [=goal], we mapped
virulence in F1 progeny derived from crosses between type II
and type III strains, which we introduced into mice [=materials
and methods]. Five virulence (VIR) loci were thus identified,
and for two of these, genetic complementation showed that a
predicted protein kinase (ROP18 and ROP16, respectively) is
the key molecule [=results]. Both are hypervariable rhoptry
proteins that are secreted into the host cell upon invasion
[=more results]. These results suggest that secreted kinases
unique to the Apicomplexa are crucial in the host-pathogen
interaction [=conclusion].”
Saeij et al., (2006) Science 314:1780-1783
• Big picture first
• Describe current state of knowledge
• Cite strategically, but be sure to include the
competition’s (= potential reviewers’) work!
• Don’t discuss or anticipate results
• End with just two sentences summarizing
what you did and what you found
• Remember, you’re telling a story!
Example End of Intro
“To determine how different strains of Toxoplasma
cause different disease, we mapped genes
responsible for virulence in mice. We show that a
family of polymorphic protein kinases are injected into
a host cell during invasion and that these play a
major role in strain-specific differences in virulence.”
• Don’t use long strings of nouns
– [“Toxoplasma-infected human foreskin fibroblast
monolayers were incubated with…”]
• Hyphenate nouns used to modify
– [“Toxoplasma-infected cells”; “parasite-containing
cells”; “gene-targeting”; etc.] [But NOT “well
established”] [Note: “we studied genes that control
the cell cycle” and “we studied cell-cycle genes”.]
• Learn to use punctuation well (it MATTERS!)
• For example, commas go at natural pauses
(as you would in speaking)
The genes, that control this
response, were identified,
without resorting to
The genes that control this
response were identified
without resorting to
Penguin press
Materials and Methods
• Goals:
– enable others to repeat your experiments
– enable the reader to understand what you did, especially
around KEY details
• Do NOT describe standard techniques that could be
done many ways and wouldn’t alter the results
– how you introduced a transgene probably isn’t important;
exactly what you introduced is
• So, DO describe crucial details of:
– cloned genes (via primers is often easiest)
– antibodies used (how raised or catalogue number and
– all strain details
– database versions, etc.
Every paragraph is a mini-paper:
• Introduction – background and context (why you’re
doing it)
• Materials and methods – what was done
• Results – what was seen
• Interpretation/conclusion – what it means (first-order
• First person OK but use sparingly and strategically:
– Unnecessary: “We transfected the parasites with…”;
preferred: “the parasites were transfected with…”
– OK: “We favor…”, “We conclude”; i.e., wherever it’s a
personal action
Tips for Results
• Begin the first or second sentence of most or even
every paragraph with “To…”
• Middle is, “The results (fig. 3) show…[literal
description of what is seen]”
• End with, “These results suggest that…”
• Use past tense in describing what was done, “To
…., the parasites were transfected with…”
• Use present tense in stating what we learn, “These
results indicate that a mutation in xxx is responsible
for …” OR conclusions that have been well vetted
previously and are generally accepted in the field.
“To…” examples
Anderson, et al. (2009) Eukaryotic Cell 8:398-409
Tips for Results
• Phrases to avoid:
– “In order to…” Use “To…”
– “Whether or not…” Just use “Whether…”
– “Since…” [“Since the parasites were transfected with a Creexpressing plasmid…” – Use “As” or “Because” to avoid
confusion about whether “since” refers to time or causality]
– “The data was were used to…”
• Hierarchy of strength for last sentences…:
“The results prove/show/demonstrate > indicate > strongly suggest
> argue for > suggest > support the notion > are consistent with
> are not inconsistent with …
General Tips
• Make it flow
• Say what you mean with the simplest words
• Avoid jargon: e.g., “the gel was run”, “the
oligonucleotide was kinased”, “the sequence was
blasted against”
• Avoid needless words (but don’t be shy about using the
ones you need!)
More Tips
• Don’t use subjective adjectives without quantitation:
– The mutants were much less virulent…
– The mutants were much less virulent (xxxx had an LD50 20-fold
higher than wild type)…
• Avoid emotive adjectives:
– Incredibly, amazingly, astonishingly similar…
• But do key in reader to whether result is expected or not:
As expected,
Consistent with the results described above,
Figure Legends
• Title should be declarative of
experiment being presented or, if
possible, (better) the point being made.
• Then describe what was done and what
is shown
• Don’t discuss the results or give any
interpretation (beyond what is in the
title; i.e., NO “these results show…”)
Figure Legends
• EVERY detail of figure must be explained:
– What do all symbols (e.g., arrows, boxing,
hatching, triangles, etc.) represent?
• Give only enough experimental detail to
understand the essence of what was done:
– don’t put in detail that could be in the materials
and methods except where there is ambiguity
• Use “Details as in Fig. 2” wherever truly
• Use big fonts so legible when reduced
• Label intuitively so reader can
understand without having to even read
the figure legend
• But still label each part “a”, “b”, etc.
• Label gel lanes by description at top
and by number at bottom
• Use arrows to highlight key aspects
• Think about whether color really helps
(often it doesn’t) – at some journals it is
• Scale to journal column size, if poss.
• Start with one or two sentence (ONLY)
of overall recap, “The results presented
here show that…”
• Use to discuss second-order context of
the work (what it all means,
implications, etc.)
• Do not go back over all the results,
only discuss those that truly need
further discussion
Discussion (cont’d)
• Share your wisdom with the reader but
don’t abandon rigor or feed the dogma
– “The results strongly suggest… We cannot
exclude the possibility, however, that …”
• Let the data speak for themselves – avoid
hyperbole and claims of “This is the
first…” [save that for the cover letter!]
• Be inclusive (think hard about this)
• Be generous (“cast your bread upon the
• Be careful about acknowledging
grant/fellowship support
• Can acknowledge grant support of a PI
who is not an author (if, e.g., a post-doc is
an author but that person’s PI is not
because s/he was really not involved)
Concluding Comments
• Write at peak times and in solid chunks
(remember just how important this is!)
• Proof-read (and spell-check) before you
pass it to anyone else. Twice, at least!
• Have a lab-mate read and micro-edit it
• Have someone outside your field read it
• Enjoy telling a story and communicating
the fruits of your labor
And lastly…
• Write assuming that ten years from now
the most aggressive, obnoxious trial
lawyer will wave your paper in your face
and challenge you on every word of
every sentence!
Thanks to my group, who happily put up with
my obsessive-compulsive editing!