ABBREVIATIONS AND CONVENTIONS
Verb Tense
ASM strongly recommends that for clarity you use the past tense to narrate particular events in the past, including the procedures, observations, and data of the study that you are reporting. Use the present tense for your own general conclusions, the conclusions of previous researchers, and generally accepted facts. Thus, most of the abstract, Materials and Methods, and Results will be in the past tense, and most of the introduction and some of the Discussion will be in the present tense.
Be aware that it may be necessary to vary the tense in a single sentence. For example, it is correct to say “White (30) demonstrated that XYZ cells grow at pH 6.8,” “Figure 2 shows that ABC cells failed to grow at room temperature,” and “Air was removed from the chamber and the mice died, which proves that mice require air.” In reporting statistics and calculations, it is correct to say “The values for the ABC cells are statistically significant, indicating that the drug inhibited. . . .”
For an in-depth discussion of tense in scientific writing, see How To Write and Publish a Scientific Paper, 7th ed.
Abbreviations
General. Abbreviations should be used as an aid to the reader, rather than as a convenience to the author, and therefore their use should be limited. Abbreviations other than those recommended by the IUPAC-IUB (Biochemical Nomenclature and Related Documents, 1992) should be used only when a case can be made for necessity, such as in tables and figures.
It is often possible to use pronouns or to paraphrase a long word after its first use (e.g., “the drug” or “the substrate”). Standard chemical symbols and trivial names or their symbols (folate, Ala, and Leu, etc.) may also be used.
Define each abbreviation and introduce it in parentheses the first time it is used; e.g., “cultures were grown in Eagle minimal essential medium (MEM).” Generally, eliminate abbreviations that are not used at least three times in the text (including tables and figure legends).
Not requiring introduction. In addition to abbreviations for Système International d’Unités (SI) units of measurement, other common units (e.g., bp, kb, and Da), and chemical symbols for the elements, the following should be used without definition in the title, abstract, text, figure legends, and tables:
- DNA (deoxyribonucleic acid)
- cDNA (complementary DNA)
- RNA (ribonucleic acid)
- cRNA (complementary RNA)
- RNase (ribonuclease)
- DNase (deoxyribonuclease)
- rRNA (ribosomal RNA)
- mRNA (messenger RNA)
- tRNA (transfer RNA)
- AMP, ADP, ATP, dAMP, ddATP, and GTP, etc. (for the respective 5′ phosphates of adenosine and other nucleosides) (add 2′-, 3′-, or 5′- when needed for contrast)
- ATPase and dGTPase, etc. (adenosine triphosphatase and deoxyguanosine triphosphatase, etc.)
- NAD (nicotinamide adenine dinucleotide)
- NAD+ (nicotinamide adenine dinucleotide, oxidized)
- NADH (nicotinamide adenine dinucleotide, reduced)
- NADP (nicotinamide adenine dinucleotide phosphate)
- NADPH (nicotinamide adenine dinucleotide phosphate, reduced)
- NADP+ (nicotinamide adenine dinucleotide phosphate, oxidized)
- poly(A) and poly(dT), etc. (polyadenylic acid and polydeoxythymidylic acid, etc.)
- oligo(dT), etc. (oligodeoxythymidylic acid, etc.)
- UV (ultraviolet)
- PFU (plaque-forming units)
- CFU (colony-forming units)
- MIC (minimal inhibitory concentration)
- Tris (tris[hydroxymethyl]aminomethane)
- DEAE (diethylaminoethyl)
- EDTA (ethylenediaminetetraacetic acid)
- EGTA (ethylene glycol-bis[β-aminoethyl ether]-N,N,N′,N′-tetraacetic acid)
- HEPES (N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid)
- PCR (polymerase chain reaction)
- AIDS (acquired immunodeficiency syndrome)
Abbreviations for cell lines (e.g., HeLa) also need not be defined.
The following abbreviations should be used without definition in tables:
- amt (amount)
- approx (approximately)
- avg (average)
- concn (concentration)
- diam (diameter)
- expt (experiment)
- exptl (experimental)
- ht (height)
- mo (month)
- mol wt (molecular weight)
- no. (number)
- prepn (preparation)
- SD (standard deviation)
- SE (standard error)
- SEM (standard error of the mean)
- sp act (specific activity)
- sp gr (specific gravity)
- temp (temperature)
- vol (volume)
- vs (versus)
- wk (week)
- wt (weight)
- yr (year)
Reporting Numerical Data
Standard metric units are used for reporting length, weight, and volume. For these units and for molarity, use the prefixes m, μ, n, and p for 10−3, 10−6, 10−9, and 10−12, respectively. Likewise, use the prefix k for 103. Avoid compound prefixes such as mµ or µµ. Use µg/ml or µg/g in place of the ambiguous ppm. Units of temperature are presented as follows: 37°C or 324 K.
When fractions are used to express units such as enzymatic activities, it is preferable to use whole units, such as “g” or “min,” in the denominator instead of fractional or multiple units, such as µg or 10 min. For example, “pmol/min” is preferable to “nmol/10 min,” and “μmol/g” is preferable to “nmol/μg.” It is also preferable that an unambiguous form, such as exponential notation, be used; for example, “μmol g−1 min−1” is preferable to “μmol/g/min.” Always report numerical data in the applicable SI units.
Representation of data as accurate to more than two significant figures must be justified by presentation of appropriate statistical analyses.
For a review of some common errors associated with statistical analyses and reports, plus guidelines on how to avoid them, see the articles by Olsen (Infect Immun 71:6689–6692, 2003; Infect Immun 82:916–920, 2014).
Statistics
Statistical analysis of data is a crucial component of scientific publication. Authors who are unsure of proper statistical analysis should have their manuscripts checked by a qualified statistician.
The following is a list of important items that must be considered before manuscript submission. Deficiencies in any of these areas may delay review and/or publication.
(i) Statistical analyses should be performed on all quantitative data regardless of how significant the differences look in the tables or figures.
(ii) Data should be appropriately analyzed as parametric (normally distributed) or nonparametric data.
(iii) Parametric and nonparametric data should be presented appropriately. Means and standard deviations or standard errors are appropriate ways of presenting data analyzed by parametric analyses (i.e., t test and analysis of variance [ANOVA]), but only medians and surrounding levels (quartiles, quintiles, and 10th and 90th percentiles, etc.) are appropriate for nonparametric statistics (Mann-Whitney test and Kruskal-Wallis test, etc.). Means have no meaning in nonparametric analyses.
(iv) For any data in which there are more than two comparisons (i.e., between one control and more than one experimental group), an analysis must be done for multigroup comparisons. Such an analysis would usually be an ANOVA for parametric data or a Kruskal-Wallis test for nonparametric data. t tests cannot be used when more than two groups are being compared (except as indicated below). Failure to use multigroup tests generates type 1 errors: concluding that two data sets within the overall data set being compared are different when in fact they are not. Exception: some statisticians argue that two-group comparisons can be used on multigroup data if the expected outcomes are appropriately anticipated before the experiment. For example, data generated by individually testing two unrelated factors for their effects on a target with only a single, untreated target as a control could be appropriately analyzed by t tests instead of ANOVA.
(v) For all appropriate multigroup comparisons, two P values must be generated and provided in the manuscript. The main P value applies to the overall data set and indicates that within that data set at least two groups differ from each other. The overall P value does not indicate which two groups are different. The main P value and the overall P value should be computed by using a post hoc test. For ANOVA, these post hoc tests are usually Dunnett's test (used to compare multiple experimental groups to a single control), the Fisher protected least significant difference (PLSD) test, the Tukey-Kramer test, and the Games-Howell test. Others may be used. Note that each post hoc test has certain underlying assumptions that may not be applicable to the data under analysis. For a Kruskal-Wallis nonparametric ANOVA, the Dunn procedure is appropriate to generate P values for two-group comparisons.
(vi) Data presented as endpoints (i.e., LD50 and ID50, etc.) contain both the calculated value and a confidence interval with a statistical significance associated with it (95%, 99%, or similar confidence interval), calculated by logit or probit analysis. Simple LD50 values, such as Reed-Muench calculations, may not be used alone.
(vii) When samples are taken multiple times from one experimental entity (i.e., multiple serum samples from one animal, gross pathology scores measured for the same animal over time, and growth curves, etc.), one cannot use analyses such as t tests, ANOVA, and the Mann-Whitney test, etc., because these tests assume that each measure is independent. An entity with a high score on day 1 is more likely to have a high score on day 2 than is an entity with a low score. It is likely that some expert statistical help will be needed for these situations, usually involving regression analysis or survival analysis, etc.
(viii) Statistical significance and biological significance are not the same. There is nothing magical about a P value of 0.05. When results from large sample sizes are compared, a P value of <0.05 will often be obtained, as P value is a function of both sample size and effect size. If sample sizes are large, then more-rigorous (i.e., smaller) P values may be desirable. If sample sizes are small, P values of >0.05 may still be important. There should be both statistical and biological significance to the results and conclusions in the manuscript.
For a review of some common errors associated with statistical analyses and reports, plus guidelines on how to avoid them, see the articles by Olsen (Infect Immun 71:6689–6692, 2003; Infect Immun 82:916–920, 2014).
Isotopically Labeled Compounds
For simple molecules, labeling is indicated in the chemical formula (e.g., 14CO2, 3H2O, and H235SO4). Brackets are not used when the isotopic symbol is attached to the name of a compound that in its natural state does not contain the element (e.g., 32S-ATP) or to a word which is not a specific chemical name (e.g., 131I-labeled protein, 14C-amino acids, and 3H-ligands).
For specific chemicals, the symbol for the isotope is placed in square brackets directly preceding the part of the name that describes the labeled entity. Note that configuration symbols and modifiers precede the isotopic symbol. The following examples illustrate correct usage.
- [14C]urea
- L-[methyl-14C]methionine
- [2,3-3H]serine
- [α-14C]lysine
- UDP-[U-14C]glucose
- E. coli [32P]DNA
- fructose 1,6-[1-32P]bisphosphate
- [γ-32P]ATP