In davemcg/SeeGEM: Integrate 'GEMINI' (and 'peddy') Output into Interactive Html Document

opts_chunk$set(echo = F,
               warning = F,
               message = F)

# decorate data, or return as given
if (params$decorate){
  document_data <- See_GEM_formatter(GEMINI_data)
} else{
  document_data <- GEMINI_data}

r params$sample_name {.tabset}

GEMINI Variation

DT::datatable(document_data$GEMINI_data, 
              width = params$table_width, 
              escape = FALSE,
              rownames = F, 
              class='compact', 
              filter='bottom',
              extensions = c('Buttons','FixedHeader', 'ColReorder'),
              options = list(columnDefs=list(list(targets=document_data$neg_core_index-1, visible = FALSE)),
                             pageLength = 15,
                             scrollX = TRUE,
                             fixedHeader = TRUE,
                             colReorder = TRUE,
                             lengthMenu = list(c(5, 15, 30, 60, 100, -1), list('5', '15', '30', '60', '100', 'All')),
                             dom = 'Bfrtip',
                             buttons = list(list(extend = 'pageLength'),
                                            list(extend = 'colvis', collectionLayout='four-column'),
                                            list(extend = 'colvisGroup', text='Core Fields', 
                                                 hide = paste(document_data$neg_core_index-1, collapse = ','), 
                                                 show = paste(document_data$core_index-1, collapse = ',')),
                                            list(extend = 'colvisGroup', text='Family Genotypes', 
                                                 hide = paste(document_data$neg_genotypes-1, collapse = ','), 
                                                 show = paste(document_data$genotypes-1, collapse = ',')),
                                            list(extend = 'colvisGroup', text='In Silico Fields', 
                                                 hide = paste(document_data$neg_in_silico_index-1, collapse = ','), 
                                                 show = paste(document_data$in_silico_index-1, collapse = ',')),
                                            list(extend='colvisGroup', text='Show All', show = ':hidden' )))) %>% 
  DT::formatStyle(columns = c('Color'), target = 'row', backgroundColor = DT::styleEqual(c(0, 1, 2), c('#ffffff', '#fff0dd', '#ffdddd'))) 

peddy QC

A small selection of outputs from the peddy tool.

For the plots, gray are all of the samples available in the vcf that peddy was run on. The colored points are samples in the family analyzed in this document.

# load peddy data
ped_check <- fread(paste0(params$peddy_path_prefix, ".ped_check.csv"))
het_check <- fread(paste0(params$peddy_path_prefix, ".het_check.csv"))
sex_check <- fread(paste0(params$peddy_path_prefix, ".sex_check.csv"))
peddy_id <- params$peddy_id

het_processor <- function(df, stat_to_filter){
  df %>% gather(stat, value, -c(sample_id)) %>% 
    filter(stat == stat_to_filter) %>% 
    mutate(value=as.numeric(value))
}
het_data <- het_processor(het_check, 'het_ratio')
median_data <- het_processor(het_check, 'median_depth')
idr_baf_data <- het_processor(het_check, 'idr_baf')

peddy Heterozygousity Check

• Median Read Depth: Median depths for the sites tested.
• Het Ratio: proportion of sites that were heterozygous.
• IDR BAF: inter-decile range (90th percentile - 10th percentile) of b-allele frequency. Large values indicate likely sample contamination.

het_ratio <- het_data %>% 
  ggplot(aes(x=stat, y=value)) + 
  geom_quasirandom(color = 'gray', alpha=0.7) + 
  geom_quasirandom(data = het_data %>% filter(sample_id %in% peddy_id), 
                   aes(x=stat,y=value, colour=sample_id)) +
  theme_minimal() + 
  xlab('') + ylab('') + ggtitle('Het Ratio') + theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())

median_depth <- median_data %>% 
  ggplot(aes(x=stat, y=value)) +
  geom_quasirandom(color = 'gray', alpha=0.7) + 
  geom_quasirandom(data = median_data %>% filter(sample_id %in% peddy_id), 
                   aes(x=stat,y=value, colour=sample_id)) +
  theme_minimal() +
  xlab('') + ylab('') + ggtitle('Median Read Depth') + theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())

idr_baf <- idr_baf_data %>% 
  ggplot(aes(x=stat, y=value)) + 
  geom_quasirandom(color = 'gray', alpha=0.7) + 
  geom_quasirandom(data = idr_baf_data %>% filter(sample_id %in% peddy_id), 
                   aes(x=stat,y=value, colour=sample_id)) +
  theme_minimal() + 
  xlab('') + ylab('') + ggtitle('IDR BAF') + theme( axis.text.x=element_blank(), axis.ticks.x=element_blank())

legend <- get_legend(idr_baf+ theme(legend.position="right"))
plot_grid(median_depth + theme(legend.position="none"),
          NULL,
          het_ratio + theme(legend.position="none"), 
          NULL,
          idr_baf + theme(legend.position="none"), 
          NULL,
          legend,
          rel_widths = c(1,0.1,1,0.1,1,0.1,0.4),
          scale=0.9,
          nrow=1) 

peddy Sex Check

• ped Sex: Sex from .ped file
• Predicted Sex: Sex predicted from rate of heterozygotes on chrX
• Error: Deviation between given and predicted sex

sex_check %>% 
  filter(sample_id %in% peddy_id) %>% 
  select('Sample ID' = sample_id, 'Pedigree Sex' = ped_sex, 'Predicted Sex' = predicted_sex, 'Error' = error) %>% 
  DT::datatable(rownames = F, 
                width = 800,
                class='compact',  
                extensions = c('Buttons'),
                options = list(dom = 'Bfrtip',
                               buttons = list(list(extend='pageLength'))))

peddy Pedigree Check

• Relatedness: the relatedness calculated from the genotypes.
• IBS2 against IBS0: IBS0 is the number of sites where the two samples shared no alleles (~0 for parent-child) and IBS2 is where the two samples were both homozygous reference, heterozygous, or homozyous alternative. This should be fairly high between related individuals. An intermediate IBS0 suggests first cousin or second cousin.

ped_check %>% 
  filter(sample_a %in% peddy_id, sample_b %in% peddy_id) %>% 
  select('Sample 1' = sample_a,
         'Sample 2' = sample_b, 
         'Relatedness' = rel,
         ibs0,
         ibs2,
         'Pedigree Parents' = pedigree_parents,
         'Predicted Parents' = predicted_parents,
         'Parent Error' = parent_error) %>% 
  DT::datatable(rownames = F, 
                width = 800,
                class='compact',  
                extensions = c('Buttons'),
                options = list(dom = 'Bfrtip',
                               buttons = list(list(extend='pageLength'))))

ped_check %>% 
  ggplot(aes(x=ibs0, y=ibs2)) + 
  geom_point(aes(x=ibs0, y=ibs2), color='gray', alpha=0.7) + 
  geom_jitter(data = ped_check %>% 
                filter(sample_a %in% peddy_id, sample_b %in% peddy_id) %>% 
                mutate('Sample - Sample' = paste0(sample_a, ' - ', sample_b)),
              aes(x=ibs0, y=ibs2, colour = `Sample - Sample`)) +
  geom_text_repel(data = ped_check %>% 
                    filter(sample_a %in% peddy_id, sample_b %in% peddy_id) %>% 
                    mutate('Sample - Sample' = paste0(sample_a, ' - ', sample_b)),
                  aes(x=ibs0, y=ibs2, label = `Sample - Sample`, colour = `Sample - Sample`),
                  box.padding = 1) + 
  theme_minimal() + 
  theme(legend.position="none")

Info

This report is auto-generated with the See GEM R package on r strftime(Sys.time(), format = "%B %d, %Y")

The analyst for this report populated the information in this dynamic documents with a variety of GEMINI queries.

FAQ:

• What does it mean if a variant is marked in red?
• This means that the variant has an impact severity of HIGH or is marked as pathogenic in ClinVar and has a gnomAD population allele frequency less than 1%.
• Orange?
• MED impact severity

WARNINGS/LIMITATIONS

• This document is only a tool to aid identification of causative variant(s)
• The in silico consequences (e.g. stop gained, synonymous) are automatically calculated, and for a variety of reasons, may not be accurate.
• The causative variants may not be present in this document

davemcg/SeeGEM documentation built on May 5, 2019, 1:34 a.m.