The Extraction of Single Nucleotide Polymorphisms and the Use of Current Sequencing Tools

1
The Extraction of Single Nucleotide Polymorphisms
and the Use of Current Sequencing Tools

• Stephen Tetreault
• Department of Mathematics and Computer Science
• Rhode Island College
• Providence, RI

2
Single Nucleotide Polymorphisms

• DNA sequence variation when a single nucleotide
  in the genome differs
• SNPs are the majority of genetic variation
• 1.4 million SNPs in a human genome
• Two haploid genomes differing at 1 SNP per 1,331
  bp
• SNPs are crucial in the effort to personalize
  medicine

3
1000 Genomes Project

• International consortium to create most complete
  catalog of human genetic variation
• Sequencing is done using utilizing next
  generation sequencing technology (e.g. Solexa,
  454, SOLiD) which is faster and less expensive
• 3 steps of the project
• Detailed scanning of six participants
• Less detailed scan of 180 participants
• Partial scans of 1000 participants

4
1000 Genomes Project

• 1000 Genomes Project Goals
• Discover genetic variants (SNPs, copy-number
  variants, indels)
• Identify frequencies of the variant alleles and
  identify their haplotype backgrounds

5
Project Focus

• Learning about the current state of sequencing
  tools
• Learning how to use these tools and understanding
  the raw data
• Creating a program to to extract the SNPs from
  the raw data and to calculate simple variant
  frequencies.
• More advanced data analysis - to be discussed in
  future works section

6
Data and Tools

• 1000 Genomes Project
• ftp//ftp-trace.ncbi.nih.gov/1000genomes/
• MAQ 0.7.1
• http//sourceforge.net/projects/maq/files/
• SAMtools 0.1.5
• http//sourceforge.net/projects/samtools/files/

7
Sequencing

• MAQ maps short reads to references and calls
  genotypes from the alignment
• MAQ maps a read to the position where the sum of
  quality values of mismatched nucleotides is
  minimum
• Issues with MAQ
• Very long run-time
• Limited computing power slowed the program down

8
Sequencing

• SAMtools was the alternative sequencing program.
• It proved faster because it could utilize BAM
  (Binary SAM) files which are prealigned partial
  scans of the participant data.
• MAQ had to align FASTA and FASTQ files, then
  change the MAP file into a Consensus file for SNP
  calling.
• SAMtools allowed for SNP calling as MAQ did
• SAMtools pileup function describes base pair
  information at each chromosomal position.

9
Sequencing

• SAMtools pileup function describes base pair
  information at each chromosomal position.

10
Project Data

• The raw data received through SAMtools pileup and
  consensus calling contains the following
  chromosome, position, reference base, consensus
  base, consensus quality score, SNP quality score,
  maximum mapping quality score, number of reads
  mapped, read bases, and base qualities.

11
Phred Quality Scores

• The consensus quality score and the SNP quality
  are Phred quality scores.
• High accuracy of Phred scores helps ensure
  reliable SNP calling

12
Finding Higher Quality SNPs

• Look at the number of reads covering the position
  with th SNP and discard those covered by three or
  fewer reads.
• Consensus quality is important, but SNP quality
  is more important. Discard a SNP with a quality
  score lower than 20.

13
A Program for Extracting SNPs

• Read in raw data line by line
• Check for SNP of high quality
• Differing reference and consensus base
• SNP with a quality score of 20 or higher
• Insert SNP as on object into array list (also
  stored in order of position)
• Keep counts for variant frequency update when
  SNP is found
• Keep count of number of SNPs per 100,000 bases
  throughout chromosome 1

14
Results

• Comparing variant frequencies
• Base change of A to G and of T to C were shown to
  be the most frequently occuring variations
• Base change of C to G was least frequently
  occuring

15
Results

• The number of SNPs occuring per 100,000 bases
  throughout chromosome 1 for participant NA07048

16
Results

• The number of SNPs occuring per 100,000 bases for
  chromosome 1 of participant NA12273. The SNPs
  appear more clustered together in frequency when
  compared to NA07048.

17
Conclusion

• Initial complications in data access and slow
  progress with MAQ were overcome.
• SAMtools proved to be faster thus more efficient
  at sequencing and SNP calling when utilizing the
  prealigned partial BAM files

18
Future Work

• FastPHASE is a program used for estimating
  missing genotypes and for reconstruction of
  haplotypes.
• Implement advanced data analysis into program by
  calling genotypes from the reads and running
  fastPHASE to obtain corresponding haplotypes.
• Look at chromosome 1 for an individual and look
  at the reads mapped covering that position and
  see what the bases are for that position to
  determine if the SNP is heterozygous or homozygous

19
Acknowledgment

• Thank you to the Professor Yufeng Wu, Jin Zhang,
  the Computer Science and Engineering Department
  at University of Connecticut, and the National
  Science Foundation for making this project and
  the Bio-Grid REU possible.