A 20,000 league view of Bioinformatics


Kelsey Florek, MPH, PhD
2019 AMD Symposium
May 23, 2019


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

https://staph-b.github.io/midwest-region/

Bioinformatics: An interdisciplinary field that develops methods and software tools for understanding biological data.

What does the data look like


						
							@M03478:141:000000000-C5B4D:1:1101:25956:10945 1:N:0:1
							TTCCGTATTCATGCAACCTATGATGAAAGTATTAGTCGGTTACTCAATGTATTTGAGCGC
							+
							ABBBBCFFFFFFGGGGGGGGGG5GHHHHHGGHHHHHHGGGGFHHHHHHHHHHHHHFGHGG
						

					

						
							@M03478:141:000000000-C5B4D:1:1101:25956:10945 1:N:0:1
							TTCCGTATTCATGCAACCTATGATGAAAGTATTAGTCGGTTACTCAATGTATTTGAGCGC
							+
							ABBBBCFFFFFFGGGGGGGGGG5GHHHHHGGHHHHHHGGGGFHHHHHHHHHHHHHFGHGG
  • M03478 - the unique instrument name
  • 141 - the run id
  • 000000000-C5B4D - flowcell id

Phred Score

  • 10: 1 in 10 90%
  • 20: 1 in 100 99%
  • 30: 1 in 1000 99.9%
  • 40: 1 in 10,000 99.99%
  • 50: 1 in 100,000 99.999%
  • 60: 1 in 1,000,000 99.9999%

what can you do with fastq / read data

k-mers: all the possible substrings of length k

https://kat.readthedocs.io

basic analysis pipeline

  • quality trimming
  • assembly
    • de novo assembly
    • reference mapping
  • antibiotic resistance detection

ensuring quality reads


						
							@M03478:141:000000000-C5B4D:1:1101:25956:10945 1:N:0:1
							TTCCGTATTCATGCAACCTATGATGAAAGTATTAGTCGGTTACTCAATGTATTTGAGCGC
							+
							ABBBBCFFFFFFGGGGGGGGGG5GHHHHHGGHHHHHHGGGGFHHHHHHHHHHHHHFGHGG
  • remove sequencing adapters
  • trim when quality drops
  • specify a minimum length
  • scan for contamination

basic analysis pipeline

  • quality trimming
  • assembly
    • de novo assembly
    • reference mapping
  • antibiotic resistance detection

de novo assembly: assembly of read data without the use of a reference sequence

de Brujin graph: a directed graph representing overlaps between sequences of symbols

de Brujin graphs

cingi sequen sfun encin cing isfu

all 4-mers: cing ingi sequ eque quen sfun enci ncin cing gisf isfu

unique 4-mers: cing ingi sequ eque quen sfun enci ncin gisf isfu

assembly graph:

sequencingisfun

difficult de Brujin graph

missis ssissi ssippi

all 4-mers: miss issi ssis ssis siss issi ssip sipp ippi

uniqe 4-mers: miss issi ssis siss ssip sipp ippi

assembly graph:

mississippi or mississississippi

choosing k

  • low k
    • more connections
    • higher chance of repeats
    • higher coverage
  • high k
    • less connections
    • higher chance of resolving repeats
    • lower coverage

storing genome assemblies
(the .fasta file)


>A/Hong_Kong/4801/2014_NP
gttaataatcactcactgagtgacatcaaagtcatggcgt
cccaaggcaccaaacggtcttatgaacagatggaaactga
tggagatcgccagaatgcaactgagattagggcatccgtc
gggaagatgattgatggaattgggagattctacatccaaa

reference mapping: a method of mapping the reads to a reference sequence

storing read mapping
(the .sam file)

  • read name / reference name
  • position read maps to on the reference sequence
  • sequence read and quality information
  • many others..

storing the read mappings in a binary format
(the .bam file)

provides a faster access to data and tends to use less memory

compression

  • gzip
  • repetitions in the data are replaced by references to the data
  • repetitions in the data are replaced by references to <7,8>
  • replaces more frequent characters with variable length encoding
  • T : 01010100 ----> T : 11

compression matters

  • Uncompressed:
    • E coli both set of reads ~900MB
    • E coli sequencing run (16 isolates) ~20GB
  • Compressed:
    • E coli both set of reads ~200MB
    • E coli sequencing run (16 isolates) ~4GB

moving data

data moves across the internet in 1,500 byte packets

  • ftp
  • http
  • sftp
  • https

basic analysis pipeline

  • quality trimming
  • assembly
    • de novo assembly
    • reference mapping
  • antibiotic resistance detection

using the data to find resistance mechanisms

  • database
    • multifasta
    • SQL
  • search for patterns

NCBI BLAST (basic local alignment search tool)

BLAST finds similar sequences by locating short matches between sequences

after the first match BLAST begins to make local alignments

location: 4377811 - 4378944

gene name: Escherichia_coli_ampC

coverage: 100

identity: 100

database: card

description: A class C ampC beta-lactamase (cephalosporinase) enzyme described in Escherichia coli shown clinically to confer resistance to penicillin-like and cephalosporin-class antibiotics.

review

  • quality control / trimming of reads
  • assembly
    • de novo
    • reference mapping
  • AR detection using BLAST

review

  • sequencing data storage
  • data compression
  • transferring data across networks

applied Linux virtual course

https://forms.gle/oKSB5KFKcv5DX4k57

Course Dates: June 10th - June 14th, 2019

Length: 2hr sessions Monday, Wednesday and Friday; Office hours on Tuesday and Thursday


kelsey.florek@slh.wisc.edu