KFlorek - WSLH Brown Bag

Advancing Data Strategies for Tracking Infectious Diseases

Kelsey Florek, PhD, MPH
Senior Genomics and Data Scientist
Wisconsin State Laboratory of Hygiene
June 6, 2024

Slides live at:
www.k-florek.net/talks

Next Generation Sequencing for Infectious Disease Surveillance

What is Next Generation Sequencing?

Sanger Sequencing

Next Generation Sequencing (Illumina)

Gillian M. Air & Heinz Schaller (1979) Rapid DNA Sequence Analysis,
CRC Critical Reviews in Biochemistry, 6:1, 1-33

High-throughput requires advanced analytics (Nextseq 2000)

30,000,000,000 ATGC's generated per sequencing run
40,000 - 150,000 words in a novel
average word length in English is 4.79
one sequencing run would generate 62,000 novels with 100,000 words each

Anatomy of a Bioinformatics Workflow


              include { SAMPLESHEET_CHECK } from '../../modules/local/samplesheet_check'

              workflow INPUT_CHECK {
                  take:
                  samplesheet // file: /path/to/samplesheet.csv
              
                  main:
                  SAMPLESHEET_CHECK ( samplesheet )
                      .csv
                      .splitCsv ( header:true, sep:',' )
                      .map { create_fasta_channel(it) }
                      .set { reads }
              
                  emit:
                  // channel: [ val(meta), [ path_to_reads ] ]
                  reads
                  // channel: [ samplesheet.valid.csv ]                   
                  csv = SAMPLESHEET_CHECK.out.csv
                  // channel: [ versions.yml ]
                  versions = SAMPLESHEET_CHECK.out.versions 
              }
              
              // Function to get list of [ meta, [ fasta ] ]
              def create_fasta_channel(LinkedHashMap row) {
                  // create meta map
                  def meta = [:]
                      meta.id         = row.sample
              
                  def fasta_meta = []
              
                  if (!file(row.fasta).exists()) {
                      exit 1
                  } else {
                      fasta_meta = [ meta, [ file(row.fasta) ] ]
                  }
              
                  return fasta_meta
              }

Sequencing Reads


              @M05192:295:000000000-K6N36:1:1101:9502:1212 1:N:0:NAGCGCTC+NCGTAAGA
              GCGTTACATGAGGCTTATACTGAAACATTGCCTAATCCCGCCCGGTGGAAAGCTAAAAANTCCTNTGAACTGCNGGGCTATTCAGAAGNNNN
              +
              CCCCCGGGGGGGGGGGGGGGGGGGGGGGGGGGGFF@@FGGGGGGGGGGGGGGDGGGGFG#:C@F#:@FGGGGG#:C@FFFGGGAFGF?####
              @M05192:295:000000000-K6N36:1:1101:13064:1213 1:N:0:NAGCGCTC+NCGTAAGA
              TAGTGGCACTGTTTGACCATCAGCAACGCATTGGTGAACTGATGCCGGAGCGGCGTTTTNACNANGCACGTCGNCAGCAAATGATGGANNNNNNNNNNNN
              +
              CCCCCGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG#:D#:#6CFGGGGG#:DFGGGFGG??FGG############
              @M05192:295:000000000-K6N36:1:1101:18301:1216 1:N:0:NAGCGCTC+NCGTAAGA
              GCCCGGTGGTGTAATTTGCGCCCTCCGAACAAAGCCACGCCACCAGGCTGGCAATCTCANACNTNGCGCCAAANCGCCGCAGAGGAATNNNNNNNN
              +
              CCCCCGGGGGFGGGGGGGGGGCFCGGGGGGGGGFGGGGGDGGEGGFGGGGGGGGGGGGF#:C#:#::@D@7FE#6CC#FGGGGCFGFE########
              @M05192:295:000000000-K6N36:1:1102:8809:23354 1:N:0:TAGCGCTC+GCGTAAGA
              GTGTTATTTGCGTGTCGCGGGCATTATGGCGGAGTACTCTCAGCCTGACGATATGATGGTGGTTTCCGCCGCCGGTAGCACCACTAACCAGTTGAT
              +
              CCC#CFGGFGGDECG9@FFGGGGFGG#,,C7::@F@FFGFFGGGGGGFGEEF@C,9EA9C@BBFGEF>FGECFGGF+F#C##BDFGFFFFE,??9,
              @M05192:295:000000000-K6N36:1:1102:12222:23342 1:N:0:TAGCGCTC+GCGTAAGA
              AAGCTAACCGATGCGGATAATGCCGCCGATGGCATTTTTTTCCCCGCCCTTGAGCAAAATATGATGGGTGCGGTGTTAATTAACGAAAATGATGAAGT
              +
              CCCCCGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGCFGGGGGGGGGGGGFGCEFGGGGGGGGGGGGGGEF
              @M05192:295:000000000-K6N36:1:1102:17996:23344 1:N:0:TAGCGCTC+GCGTAAGA
              GAGCAGGATAAAACCTACAAAATTACAGTTCTGCATACCAATGATCATCATGGGCATTTTTGGCGCAATGAATATGGCGAATATGGTCTGGCG
              +
              CCCCCGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGDGGGGGGGGGGGGGGGGGGFGGGGGG

Phylogenetic Tree

WSLH CDD Bioinformatic Workflows

Spriggan
AR Report Generator
SPNTypeID

Public Health Application of
Infectious Disease Genomics

Pathogen Surveillance
- trends
- prioritization and risk assessment
- early warning system

Outbreak Response
- source tracking
- containment
- risk assessment

Pathogen Surveillance: SARS-CoV-2

SARS-CoV-2 Wastewater

SARS-CoV-2 Clinical

Advancing Genomics in the Cloud

What is the cloud?

Cloud Data Strategies

Genomic Data Lake

Genomic Data Cloud Infrastructure

Monitoring Data Workflows with Nextflow Tower

WSLH Data Portal Infrastructure

AWS Glue and MapReduce

Map: input data is first split into smaller blocks, then each block is assigned to a separate worker node and processed in parallel
Shuffle: worker nodes redistribute data such that all data belonging to a group located on the same worker node
Reduce: if needed worker nodes then process each group of output data, in parallel