Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Rmd work #12

Merged
merged 3 commits into from
Sep 20, 2024
Merged

Rmd work #12

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
4e56a4d
additional work on cleaning up Rmds
kweav Jul 2, 2024
4830d1d
resolve conflict with main
kweav Aug 10, 2024
256638f
Merge branch 'updateDemographics' into rmdWork
kweav Sep 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
231 changes: 136 additions & 95 deletions anvilPoll2024ExtraAnalysis.Rmd
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,17 +13,18 @@ library(magrittr) #for %<>%
library(grid) #for Grobs and unit()
library(ggrepel) #for geom_text_repel()
library(patchwork)
library(ggVennDiagram)

knitr::knit_child(here("anvilPoll2024MainAnalysis.Rmd"))
```

# Supplemental Analyses
# Supplemental Analyses and Graphs

## Identify type of user
## Identify type of user (supplemental)

No supplemental analyses
*No supplements at this time*

## Demographics: Institutional Affiliation
## Demographics: Institutional Affiliation (supplemental)

### Number of institutions represented in responses

Expand Down Expand Up @@ -236,7 +237,7 @@ combined_plot
ggsave(here("plots/institutionalType_facetedUserType.png"), plot = combined_plot)
```

## Demographics: Highest Degree Attained
## Demographics: Highest Degree Attained (supplemental)

<details><summary>Description of variable definitions and steps</summary>

Expand Down Expand Up @@ -278,17 +279,15 @@ ggsave(here("plots/degree_usertype.png"))

## Demographics: Kind of Work

No supplemental analyses
*No supplements at this time*

## Demographics: Consortia Affiliations

No supplemental analyses
*No supplements at this time*

## Experience: Genomics and Clinical Research Experience
## Experience: Genomics and Clinical Research Experience (supplemental)

### Plot data

<details><summary>Should we split current users vs potential users?</summary>
### Should we split current users vs potential users?

Here we use two different plots to show that the distribution of experience level among these three research types is similar when comparing the distribution of current users vs potential users. In this first plot, we have the experience level on the x-axis, the count on the y-axis, and color the bars by research type. We stack the user type responses using `facet_wrap` and `nrow=2` as an argument within that. We use a `position="dodge"` to cluster the similar research type bars next to each other. And we use geom_text to label the bars with the actual count. This requires `group = researchType` within the `geom_text()` `aes()` and `position = position_dodge(width = 0.9)` within the general `geom_text()` function.

Expand Down Expand Up @@ -337,9 +336,60 @@ ggsave(here("plots/researchExperienceLevel_colorExperience.png"))

Both of these give us confidence that current and potential user counts for reported experience level in these research areas show similar distributions. So we'll go ahead and plot it without splitting out `UserType`.

</details>
### Overlap in experience levels for moderate or extreme experience categories for respondents

```{r}
resultsTidy %>%
select(Timestamp, HumanGenomicExperience, HumanClinicalExperience, NonHumanGenomicExperience, UserType) %>%
pivot_longer(c(HumanGenomicExperience,
HumanClinicalExperience,
NonHumanGenomicExperience),
names_to = "researchType",
values_to = "experienceLevel") %>%
mutate(experienceLevel =
factor(experienceLevel,
levels = c("Not at all experienced",
"Slightly experienced",
"Somewhat experienced",
"Moderately experienced",
"Extremely experienced")),
researchType = case_when(researchType == "HumanClinicalExperience" ~ "Human Clinical\nResearch",
researchType == "HumanGenomicExperience" ~ "Human Genomic\nResearch",
researchType == "NonHumanGenomicExperience" ~ "Non-human\nGenomic Research"),
Timestamp = factor(Timestamp)) %>%
ggplot(aes(y = factor(experienceLevel,
levels = rev(c("Not at all experienced",
"Slightly experienced",
"Somewhat experienced",
"Moderately experienced",
"Extremely experienced"))),
x = Timestamp,
fill = experienceLevel)) +
geom_tile() +
scale_fill_manual(values = c("#035C94","#035385","#024A77","#024168", "#02395B")) +
theme_bw() +
theme(axis.text.x=element_blank(),
axis.ticks.x=element_blank(),
axis.text.y = element_blank(),
axis.ticks.y = element_blank(),
legend.position = "left") +
ylab("") +
ggtitle("How much experience do you have analyzing\nthe following data categories?") +
xlab("Respondent") +
facet_wrap(~researchType, nrow=3, strip.position="right")
```

```{r}
inputList <- list(ClinicalExperience = which(resultsTidy$clinicalFlag),
HumanGenomicsExperience = which(resultsTidy$humanGenomicFlag),
NonHumanGenomicsExperience = which(resultsTidy$nonHumanGenomicFlag))

ggVennDiagram(inputList,
category.names = c("Clinical\nExperience", "Human Genomics\nExperience", " Non-human Genomics Experience")) +
scale_x_continuous(expand = expansion(mult = .2))
```

#### Preferred bar plot
### Alternate plot

<details><summary>Description of variable definitions and steps</summary>

Expand Down Expand Up @@ -372,9 +422,9 @@ ggsave(here("plots/researchExperienceLevel_noColor_noUserTypeSplit.png"))

## Experience: Controlled Access Datasets

No Supplemental Analyses
*No supplements at this time*

## Experience: Tool & Resource Knowledge/Comfort level
## Experience: Tool & Resource Knowledge/Comfort level (supplemental)

```{r}

Expand Down Expand Up @@ -465,7 +515,7 @@ ggsave(here("plots/dataresources_comfortscore.png"))



## Preferences: Feature Importance Ranking
## Preferences: Feature importance for current vs potential users (supplemental)

### Numerical response bias

Expand Down Expand Up @@ -569,7 +619,6 @@ ggplot(densitydf, aes(x=value, group = Feature, fill = Feature)) +
ggsave(here("plots/densityplot_rankfeatures.png"))
```


#### Density plot with facets for feature

<details><summary>Description of variable definitions and steps</summary>
Expand Down Expand Up @@ -678,110 +727,102 @@ ggplot(countdf, aes(fill=rank, y=Feature, x=n)) +
ggsave(here("plots/stackedbarplot_rankfeatures.png"))
```

### Plot y-axis ordered by potential user ratings

## Types of Data respondents would want to analyze on the AnVIL

<details><summary>Question and possible answers</summary>

>What types of data do you or would you analyze using the AnVIL?

Possible answers include

* Genomes/exomes
* Transcriptomes
* Metagenomes
* Proteomes
* Metabolomes
* Epigenomes
* Structural
* Single Cell
* Imaging
* Phenotypic
* Electronic Health Record
* Metadata
* Survey
* Other (with free text response)

</details>
```{r}

<details><summary>Description of variable definitions and steps</summary>
# Provide a list of AnVIL only Tools
AnVIL_only <-
setdiff(toPlotToolKnowledgeSeparateBR[toPlotToolKnowledgeSeparateBR$UserType == "Current Users" &
toPlotToolKnowledgeSeparateBR$AnVILorNo == "On the AnVIL", ]$Tool,
toPlotToolKnowledgeSeparateBR[toPlotToolKnowledgeSeparateBR$UserType == "Potential Users", ]$Tool)

Because the responses for this data are going to look very similar to the controlled access dataset question (above), we'll follow similar reasoning here in how we prepare and plot the data.
# Order dummy column based only on Potential users
toPlotToolKnowledgeSeparateBR <-
toPlotToolKnowledgeSeparateBR %>% mutate(ToolOrder = case_when(
UserType == "Potential Users" | Tool %in% AnVIL_only ~ avgScore,
TRUE ~ 0
))

</details>

### Prepare the data
PlotToolKnowledge_potential_user_score <-
ggplot(data = toPlotToolKnowledgeSeparateBR) +
geom_point(data = toPlotToolKnowledgeSeparateBR[toPlotToolKnowledgeSeparateBR$UserType == "Potential Users" | toPlotToolKnowledgeSeparateBR$Tool %in% AnVIL_only ,],
aes(color = UserType, shape = AnVILorNo, y = reorder(Tool, ToolOrder), x = avgScore)) +
geom_point(data = toPlotToolKnowledgeSeparateBR[toPlotToolKnowledgeSeparateBR$UserType == "Current Users",],
aes(color = UserType, shape = AnVILorNo, y = Tool, x = avgScore))

```{r}
prep_df_typeData <- function(subset_df){
subset_df %<>% separate(TypesOfData, c("WhichA", "WhichB", "WhichC", "WhichD", "WhichE", "WhichF", "WhichG", "WhichH", "WhichI", "WhichJ", "WhichK", "WhichM", "WhichN", "WhichO"), sep = ", ", fill="right") %>%
pivot_longer(starts_with("Which"), names_to = "WhichChoice", values_to = "whichTypeData") %>%
drop_na(whichTypeData) %>%
group_by(whichTypeData) %>% summarize(count = n()) %>%
mutate(whichTypeData =
recode(whichTypeData,
"I don't analyze data on AnVIL" = NA_character_,
"I store data in AnVIL. I don’t analyze it." = NA_character_,
"Used in training for analysis of genomes (variant calling)" = "Variant Calling"
)
) %>%
drop_na(whichTypeData)
return(subset_df)
}
PlotToolKnowledge_customization(PlotToolKnowledge_potential_user_score)
ggsave(here("plots/tooldataresourcecomfortscore_singlepanel_by_potential_users.png"), w = 2200, h = 1350, units = "px")
```

```{r}
typeOfDataDf <- resultsTidy %>% prep_df_typeData()

typeDataClinicalSubset <- resultsTidy %>%
filter(clinicalFlag == TRUE) %>%
prep_df_typeData()
### simpler plots focusing on a subset of the data

typeDataHumanGenomicSubset <- resultsTidy %>%
filter(humanGenomicFlag == TRUE) %>%
prep_df_typeData()
```
```{r}
#only separate from the AnVIL data

### Plot the data
simplerPlot <- toPlotToolKnowledge %>%
filter(AnVILorNo == "Separate from the AnVIL") %>%
ggplot(aes(y = reorder(Tool, avgScore), x=avgScore)) + geom_point(aes(color = UserType)) +
geom_line() +
scale_x_continuous(breaks = 0:5, labels = 0:5, limits = c(0,5)) + ylab("Tool or Resource") + xlab("Average Knowledge or Comfort Score") + theme_bw() + theme(panel.background = element_blank(), panel.grid.minor.x = element_blank()) +
annotation_custom(textGrob("Don't know\nat all", gp=gpar(fontsize=8, fontface = "bold")),xmin=0,xmax=0,ymin=-1,ymax=-1) +
annotation_custom(textGrob("Extremely\ncomfortable", gp=gpar(fontsize=8, fontface= "bold")),xmin=5,xmax=5,ymin=-1,ymax=-1) +
coord_cartesian(clip = "off") +
theme(plot.margin = margin(1,1,1,1.1, "cm"))+
ggtitle("How would you rate your knowledge of or\ncomfort with these technologies\n(separate from the AnVIL)?") +
theme(legend.title = element_blank())

```{r}
simplerPlot

plot_type_data <- function(inputToPlotDF, subtitle = NULL){
toreturnplot <- ggplot(inputToPlotDF, aes(x = reorder(whichTypeData, -count), y = count)) +
geom_bar(stat="identity") +
theme_classic() + theme(panel.background = element_blank(), panel.grid = element_blank()) +
theme(axis.text.x = element_text(angle=45, hjust=1)) +
xlab("Types of data") + ylab("Count") +
ggtitle("What types of data do you or would you analyze using the AnVIL?", subtitle = subtitle) +
geom_text(aes(label = after_stat(y), group = whichTypeData),
stat = 'summary', fun = sum, vjust = -1, size=2) +
coord_cartesian(clip = "off")
return(toreturnplot)
}
ggsave(here("plots/toolsSeparateFromAnVIL_comfortscore.png"), plot = simplerPlot)
```

```{r}
everyone_type_data <- plot_type_data(typeOfDataDf)
#add in purple points of comparison for On the AnVIL

everyone_type_data
toPlot_simplified <- toPlotToolKnowledge %>%
filter(AnVILorNo == "Separate from the AnVIL")

ggsave(here("plots/typesOfData.png"), plot=everyone_type_data)
onAnVIL <- toPlotToolKnowledge %>%
filter(AnVILorNo == "On the AnVIL") %>%
right_join(., toPlot_simplified,by = "Tool") %>%
bind_rows(.,
data.frame(Tool = "RStudio",
avgScore.x = toPlotToolKnowledge[which(toPlotToolKnowledge$Tool == "Bioconductor & RStudio"),"avgScore"],
UserType.x = "Current Users",
AnVILorNo.x = "On the AnVIL"),
data.frame(Tool = "Bioconductor",
avgScore.x = toPlotToolKnowledge[which(toPlotToolKnowledge$Tool == "Bioconductor & RStudio"),"avgScore"],
UserType.x = "Current Users",
AnVILorNo.x = "On the AnVIL")
) %>% drop_na(avgScore.x)
```

```{r}
clinical_type_data <- plot_type_data(typeDataClinicalSubset, subtitle = "Respondents moderately or extremely experienced with clinical data")

clinical_type_data
```{r}
simplerPlot + geom_point(data = onAnVIL, aes(x=avgScore.x,y=Tool,colour="#C77CFF")) +
scale_color_manual(
values = c("#F8766D", "#00BFC4", "#C77CFF"), labels = c("Potential Users", "Current Users", "Current User Ratings\nfor related AnVIL tools")) + theme(legend.title = element_blank())

ggsave(here("plots/typesOfData_clinical.png"), plot=clinical_type_data)
ggsave(here("plots/tools_comfortscore.png"))
```

```{r}
humangenomic_type_data <- plot_type_data(typeDataHumanGenomicSubset, subtitle = "Respondents moderately or extremely experienced with human genomic data")
#only the data resources

humangenomic_type_data
toPlotToolKnowledge %>%
filter(Tool == "DUOS" | Tool == "Access controlled access data" | Tool == "TDR" | Tool == "Terra Workspaces") %>%
ggplot(aes(y = reorder(Tool, avgScore), x=avgScore)) + geom_point(colour = "#F8766D") +
scale_x_continuous(breaks = 0:5, labels = 0:5, limits = c(0,5)) + ylab("Data Resource") + xlab("Average Knowledge or Comfort Score") + theme_bw() + theme(panel.background = element_blank(), panel.grid.minor.x = element_blank()) +
annotation_custom(textGrob("Don't know\nat all", gp=gpar(fontsize=8, fontface = "bold")),xmin=0,xmax=0,ymin=-0.35,ymax=-0.35) +
annotation_custom(textGrob("Extremely\ncomfortable", gp=gpar(fontsize=8, fontface= "bold")),xmin=5,xmax=5,ymin=-0.35,ymax=-0.35) +
coord_cartesian(clip = "off") +
theme(plot.margin = margin(1,1,1,1.1, "cm"))+
ggtitle("How would you rate your knowledge of or\ncomfort with these AnVIL data features?") +
theme(legend.title = element_blank())

ggsave(here("plots/typesOfData_humangenomic.png"), plot=humangenomic_type_data)
ggsave(here("plots/dataresources_comfortscore.png"))
```

## Source of funds for cloud computing
Expand Down
Loading