Metagenomic biomarker discovery and explanation

Metagenomic biomarker discovery and explanation Nicola Segata Jacques Izard Levi Waldron Dirk Gevers Larisa Miropolsky Wendy S. Garrett Curtis Huttenhower

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Metagenomic biomarker discovery and explanation 2011 Genome biology 10.1186/gb-2011-12-6-r60 segata2011-metagenomic-biomarker-discovery-an segata2011-metagenomic-biomarker-discovery-an.txt, the project's staged copy d28336cb46623b411f1f8d70aa4ef80e247cb155 cc-by biology

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them metagenomic features - organisms, clades, OTUs, genes or pathways - that differ between two or more microbial communities Quotation withheld because this record is not cleared for redistribution. how to find biomarkers that EXPLAIN a difference between conditions rather than merely differ across them Quotation withheld because this record is not cleared for redistribution. LEfSe: significance, biological consistency and effect size coupled into one procedure, which the paper says no existing method provides together Quotation withheld because this record is not cleared for redistribution. method Quotation withheld because this record is not cleared for redistribution. differential abundance of microbial features between phenotypes Quotation withheld because this record is not cleared for redistribution. system-or-tool-building system-or-tool-building Quotation withheld because this record is not cleared for redistribution. mechanisms THREE TESTS IN SEQUENCE, EACH ASKING A DIFFERENT QUESTION: a Kruskal-Wallis rank test for whether the feature differs at all, pairwise Wilcoxon tests among SUBCLASSES for whether the difference is consistent with the population structure, and LDA for how big it is about this quote: the quote carries a line-break hyphen scar: the paper's “Kruskal-Wallis” is staged as “Kruskal- Wallis”. Verbatim against the staged text; the damage is mid-name and cannot be re-sliced around. Confidence 0.85 under an earlier note of ours. Quotation withheld because this record is not cleared for redistribution. mechanisms EXPLAINING RATHER THAN DETECTING: the method keeps features able to explain the difference between conditions, not merely those with uneven distributions - which is what a significance test alone returns Quotation withheld because this record is not cleared for redistribution. mechanisms PRIOR KNOWLEDGE AS A CONSTRAINT ON A HIGH-DIMENSIONAL SEARCH: the biological hypothesis is encoded and used to narrow the space, which is the paper's answer to having far more candidate biomarkers than samples Quotation withheld because this record is not cleared for redistribution. THE TRADE IS STATED AND PRICED: a substantially better false positive rate is bought with a moderately increased false negative rate, and the paper says the user can adjust where it sits Quotation withheld because this record is not cleared for redistribution. There are typically far more candidate biomarkers than samples, which is a property of the data and not of the method about this quote: the quote carries a spurious space inside a word (taxonomy shape 8): “associated with” is staged as “associated w ith”. Confidence 0.85. Quotation withheld because this record is not cleared for redistribution. Microbial communities show remarkably high inter-subject variability, alongside sequencing errors and chimeric reads Quotation withheld because this record is not cleared for redistribution. Finding a biomarker is stated to be only a PRECURSOR to understanding a mechanism - the method does not claim to deliver the biology Quotation withheld because this record is not cleared for redistribution. the approach validated on three independent bodies of data - human microbiomes, a mouse model of ulcerative colitis, and environmental samples eliminates: a biomarker method tuned to one dataset's structure - three data sources of different kinds are what shows the discovery is not a property of any one of them · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. class comparison class comparison Quotation withheld because this record is not cleared for redistribution. class discovery class discovery Quotation withheld because this record is not cleared for redistribution. class prediction class prediction Quotation withheld because this record is not cleared for redistribution. LEfSe LEfSe Quotation withheld because this record is not cleared for redistribution. effect size effect size Quotation withheld because this record is not cleared for redistribution. biological consistency biological consistency Quotation withheld because this record is not cleared for redistribution. effect relevance effect relevance Quotation withheld because this record is not cleared for redistribution. statistical significance statistical significance Quotation withheld because this record is not cleared for redistribution. biomarkers biomarkers Quotation withheld because this record is not cleared for redistribution. false positive rate false positive rate Quotation withheld because this record is not cleared for redistribution. false negative rate false negative rate Quotation withheld because this record is not cleared for redistribution. P-value P-value Quotation withheld because this record is not cleared for redistribution. inter-subject variability inter-subject variability Quotation withheld because this record is not cleared for redistribution. Wilcoxon rank-sum test Wilcoxon rank-sum test Quotation withheld because this record is not cleared for redistribution. prior biological knowledge prior biological knowledge Quotation withheld because this record is not cleared for redistribution. high-dimensional data high-dimensional data Quotation withheld because this record is not cleared for redistribution. class comparison - EXPLAINING a difference between phenotypes in terms of consistent pathways or mechanisms. Distinguished by the paper from class DISCOVERY (finding subtypes) and class PREDICTION (assigning a new sample) - three tasks, one dataset, and the paper says which one it is doing Quotation withheld because this record is not cleared for redistribution. effect size - USED HERE AS AN ORTHOGONAL AXIS TO THE P-VALUE, not as a refinement of it: a very significant difference can be too small to matter, and a marker at p = 0.01 can carry a large effect whose significance was eaten by technical noise Quotation withheld because this record is not cleared for redistribution. class comparison —distinct-from→ class prediction class comparison class prediction strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. effect size —distinct-from→ statistical significance effect size statistical significance strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. biological consistency —part-of→ LEfSe biological consistency LEfSe strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution. effect relevance —part-of→ LEfSe effect relevance LEfSe strength of the relation: 1 Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1186/gb-2011-12-6-r60.