In brouwern/mammalsmilkRA: What the Package Does (One Line, Title Case)

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  warning = F,
  message = F
)

Introduction

A diagram is often very useful for showing where a study took place, how an experiment was designed, how treatments were allocated, or how there are dependencies in the data. Zuur & Ieno (2016) emphasize this ("Step 2: Visualize the experimental design" and "Step 4: Identify the dependency structure in the data"; see Figure 4).

What follows are references to different types of diagrams that you could include in your study.

Below is also an example using a graph of the hierarchical phylogenetic relationships.

Site maps: Where a study took place

These papers have examples of site maps for ecological studies; both maps are in the supplemental information. The focus in more on providing contextual geographic information than the details of the study.

Latta, Brouwer et al. Avian community characteristics and demographics reveal how conservation value of regenerating tropical dry forest changes with forest age. PeerJ https://peerj.com/articles/5217/#supp-1

Latta, Brouwer et al. Long-term monitoring reveals an avian species credit in secondary forest patches of Costa Rica. PeerJ https://peerj.com/articles/3539/#supplemental-information

Experimental layout maps

Ecological studies sometimes include maps to show the geometric arrangement of study plots or experimental treatments.

Historic disturbance regimes promote tree diversity only under low browsing regimes in eastern deciduous forest T Nuttle, AA Royo, MB Adams… - Ecological Monographs, 2013 - Wiley Online Library

Long-term biological legacies of herbivore density in a landscape‐scale experiment: forest understoreys reflect past deer density treatments for at least 20 years T Nuttle, TE Ristau, AA Royo - Journal of Ecology, 2014

Accounting for multilevel data structures in fisheries data using mixed models T Wagner, DB Hayes, MT Bremigan - Fisheries, 2006

Dependency structure diagrams

For complex experiments it can be useful to create a diagram that illustrate the hierarchical structure of the data, such as students in class rooms in schools in school districts in counties in states.

MCDB / Bio-medicine

Improving basic and translational science by accounting for litter-to-litter variation in animal models SE Lazic, L Essioux - BMC neuroscience, 2013

Four simple ways to increase power without increasing the sample size Stanley E Lazic Laboratory Animals https://journals.sagepub.com/doi/full/10.1177/0023677218767478

Analyzing clustered data: why and how to account for multiple observations nested within a study participant? EL Moen, CJ Fricano-Kugler, BW Luikart, AJ O'Malley - Plos one, 2016

Ecology

Nested by design: model fitting and interpretation in a mixed model era H Schielzeth, S Nakagawa - Methods in Ecology and Evolution, 2013

Meta-evaluation of meta-analysis: ten appraisal questions for biologists S Nakagawa, DWA Noble, AM Senior… - BMC …, 2017 https://bmcbiol.biomedcentral.com/articles/10.1186/s12915-017-0357-7

Nonindependence and sensitivity analyses in ecological and evolutionary meta-analyses …, M Lagisz, RE O'dea, S Nakagawa - Molecular …, 2017

Zuur & Ieno. 2016. A protocol for conducting and presenting results of regression‐type analyses https://besjournals.onlinelibrary.wiley.com/doi/full/10.1111/2041-210X.12577

Graphical abstracts

Graphical abstracts are becoming increasingly common in some fields.

https://www.elsevier.com/authors/journal-authors/graphical-abstract

http://crosstalk.cell.com/blog/attract-readers-at-a-glance-with-your-graphical-abstract

https://bitesizebio.com/31125/how-to-make-a-good-graphical-abstract/

Flow charts

Clinical trials

Clinical trials often have a flow chart showing how an initial patient population was screen and then allocated to the experimental groups.

https://openi.nlm.nih.gov/detailedresult.php?img=PMC2945952_1471-2474-11-205-2&req=4

Meta-analyses

Meta-analyses often have flow charts showing how an initial pool of studies was screen for inclusion in the analysis. The PRISMA guidlines (Preferred Reporting Items for Systematic Reviews & Meta-Analyses) recommend a 4-phase flow diagram be included in meta-analyses to describe the process of finding, screening, and extracting data from papers in a meta analysis. See Libarati et al 2009

I've copied an example using You'll need to install the package if you want to make it yourself eg run install.packages("PRISMAstatement"). See the PRISMAstatement package for details

library(PRISMAstatement)
prisma(found = 750,
       found_other = 123,
       no_dupes = 776, 
       screened = 776, 
       screen_exclusions = 13, 
       full_text = 763,
       full_text_exclusions = 17, 
       qualitative = 746, 
       quantitative = 319,
       width = 300, height = 300)

A real example is at https://openi.nlm.nih.gov/detailedresult.php?img=PMC2559819_1471-2261-8-24-1&req=4

Coordinated experiments across labs

The paper below has an interesting flow chart about how samples were shared across three labs to test for biases in lab techniques. The ends result was a retraction of a Cell paper!

Resolving discrepant findings on ANGPTL8 in beta-cell proliferation: a collaborative approach to resolving the betatrophin controversy AR Cox, O Barrandon, EP Cai, JS Rios, J Chavez… - PloS one, 2016

Dependency structure due to phylogeny

The dependency structure in the mammals milk data is due to phylogeny. The original paper presents a phylogenetic tree as a figure. Similarly, code below is a first attempt to visualize this based on nominal taxonomic categories (class, order, family, genus, species). Its not very clean yet.

Conceptual issues of clustering

The following papers discuss conceptual issues of grouping, clustering etc, but don't have pictures of it.

A solution to dependency: using multilevel analysis to accommodate nested data E Aarts, M Verhage, JV Veenvliet, CV Dolan… - … neuroscience, 2014

What exactly is 'N'in cell culture and animal experiments? SE Lazic, CJ Clarke-Williams, MR Munafò - PLoS biology, 2018

The problem of pseudoreplication in neuroscientific studies: is it affecting your analysis? SE Lazic - BMC neuroscience, 2010

Using biological insight and pragmatism when thinking about pseudoreplication N Colegrave, GD Ruxton - Trends in ecology & evolution, 2017

Visualizing the structure of the mammals milk data

What follows is a code dump to make a diagram of the mammals milk data set. Its not very clean yet; scroll down a ways to the bottom to see the output.

Preliminaries

Load packages

library(data.tree)
library(treemap)
library(here)
library(stringr)

Load data

Load data in R

This is particular to my computer

file. <- "Appendix-2-Analysis-Data_mammalsmilkRA.csv"
path. <- here("/inst/extdata",file.)

milk <- read.csv(path., skip = 3)


milk$ord <- gsub("11","",milk$ord)

Cleaning

genus_spp_subspp_mat <- milk$spp %>% 
         str_split_fixed(" ", n = 3)

milk$genus <- genus_spp_subspp_mat[,1]

Set up taxonomy

Class, subclass

## class - all mammals
milk$class <- "Mammalia"

## Subclass

### everyone but Monotrmes are Theriiformes
#### = Yinotheria
milk$subclass <- "Theriiformes"
milk$subclass[which(milk$ord == "Monotremata")] <- "Yinotheria"

Cohort

milk$cohort <- "Placentalia"


Marsupialia <- c("Diprotodontia"  #marsupials
                      ,"Didelphimorphia" #opossums
                      ,"Dasyuromorphia"#tasmanian devil)
                      ,"Peramelemorphia" #bandicoots
                  )

milk$cohort[which(milk$ord %in% Marsupialia)] <- "Marsupialia"


milk$cohort[which(milk$ord == "Monotremata")] <- "Australosphenida"

magnorder

milk$magnorder <- "Xenarthra-or-Epitheria"

milk$magnorder[which(milk$ord == "Monotremata")] <- "Monotremata-magnorder"

 Australidelphia <- c("Diprotodontia"  #marsupials
                      #,"Didelphimorphia" #opossums
                      ,"Dasyuromorphia"#tasmanian devil)
                      ,"Peramelemorphia" #bandicoots
                  )



milk$magnorder[which(milk$ord %in% Australidelphia)] <- "Australidelphia"

milk$magnorder[which(milk$ord == "Didelphimorphia")] <- "Ameridelphia"

Use tables to look at dependencies

with(milk,
     table(class, subclass))

with(milk,
     table(subclass, cohort))
with(milk,
     table(cohort, magnorder))
with(milk,
     table(ord, magnorder))

Make hierachical graph

Set up path string

#milk$i <- 1:nrow(milk)
milk$pathString <- with(milk,
                        paste(class,  #Mammalia
                              subclass,
                              cohort,
                              magnorder,
                              ord,
                              sep = "/"))

Set up nodes

milk_tree <- as.Node(milk)

Print hiearchy as text

print(milk_tree, limit = 20)

Plot

Plot branch diagram

plot(milk_tree)

Plot dendrogram

plot(as.dendrogram(milk_tree),
     horiz = T,
     center = T, 
     cex = 0.5,
     nodePar = 1)

brouwern/mammalsmilkRA documentation built on May 3, 2019, 7:39 p.m.