Data Basics. Josef K Vogt Slides by: Simon Rasmussen Next Generation Sequencing Analysis

Transcription

1 Data Basics Josef K Vogt Slides by: Simon Rasmussen 2017

2 Generalized NGS analysis Sample prep & Sequencing Data size Main data reductive steps SNPs, genes, regions Application Assembly: Compare Raw Pre- specific: Question Alignment / samples / Answer? reads processing Variant calling, de novo methods count matrix,...

3 What is sequence data? Sequences are stored in fasta-files Header Sequence >gi ref NC_ Escherichia coli chromosome, complete genome GTAAGTATTTTTCAGCTTTTCATTCTGACTGCAACGGGCAATATGTCTCTGTGTGGATTAAAAAAAGAGT GTCTGATAGCAGCTTCTGAACTGGTTACCTGCCGTGAGTAAATTAAAATTTTATTGACTTAGGTCACTAA ATACTTTAACCAATATAGGCATAGCGCACAGACAGATAAAAATTACAGAGTACACAACATCCATGAAACG CATTAGCACCACCATTACCACCACCATCACCATTACCACAGGTAACGGTGCGGGCTGACGCGTACAGGAA ACACAGAAAAAAGCCCGCACCTGACAGTGCGGGCTTTTTTTTCGACCAAAGGTAACGAGGTAACAACCAT GCGAGTGTTGAAGTTCGGCGGTACATCAGTGGCAAATGCAGAACGTTTTCTGCGTGTTGCCGATATTCTG GAAAGCAATGCCAGGCAGGGGCAGGTGGCCACCGTCCTCTCTGCCCCCGCCAAAATCACCAACCACCTGG TGGCGATGATTGAAAAAACCATTAGCGGCCAGGATGCTTTACCCAATATCAGCGATGCCGAACGTATTTT TGCCGAACTTTTGACGGGACTCGCCGCCGCCCAGCCGGGGTTCCCGCTGGCGCAATTGAAAACTTTCGTC GATCAGGAATTTGCCCAAATAAAACATGTCCTGCATGGCATTAGTTTGTTGGGGCAGTGCCCGGATAGCA E.coli ~ Mbases Human ~ 3.2 Gbases

4 Then what is NGS data? Fastq Sequence ATTCCCGGCCTTTTTCCAGGCCTGCCTGCTCGAGC + BAAAGECEE<EEDFEDF3DBDBB=A+==>9>>88? Qualities (prob. that base call is wrong) Millions to billions of these

5 Quality score encoding Quality score is the combination of these two (Illumina): Quality predictor values of clusters: Intensity profiles and signal-to-noise ratios Quality model/table: Pre-calculated combinations of the above Depend on machine, chemistry, software

6 A closer look at the qualities Sequence ATTCCCGGCCTTTTTCCAGGCCTGCCTGCTCGAGC + BAAAGECEE<EEDFEDF3DBDBB=A+==>9>>88? Qualities (prob. that base call is wrong) One character encodes a number Phred-scale using ascii table (0-255) Q = -l0 * log10 P This number (Q) can be converted to P P = 10^(-Q/10)

7 Phred ATTCCCGGCCTTTTTCCAGGCCTGCCTGCTCGAGC + BAAAGECEE<EEDFEDF3DBDBB=A+==>9>>88? ~1e-6 Q ~ Prob 10 ~ ~ ~ ~

8 Phred-scaled probabilities Base qualities, read mapping qualities, variant qualities,... Straight-forward, except for when they are used in reads! Offset: Sanger = 33, Illumina = ATTCCCGGCCTTTTTCCAGGCCTGCCTGCTCGAGC + BAAAGECEE<EEDFEDF3DBDBB=A+==>9>>88? Q ~ Prob Phred: ~1e-6 Sanger: ~0.001 HUGE difference! 10 ~ ~ 0.01 Illumina: ~1 Exercise today 30 ~ ~

9 Sanger vs. Illumina vs. Solexa 454, Ion Torrent, Pac Bio, Sanger: Sanger encoding Illumina reads: Illumina or Sanger encoding. New reads are all Sanger Solexa data: Solexa encoding (bought by Illumina) All data from SRA: Sanger

10 Read types Fragment DNA: Single end Paired end Ins: bp Mate pair Ins: 2kb - 40kb (~5kb) Protocol/technology dependent

11 Read orientation Single end Forward Paired end Illumina: Forward - Reverse Mate pair Illumina: Reverse - Forward Different for other technologies!

12 Single end reads: Mate pairs: Special applications Sometimes the only possibility (small DNA fragments / ancient DNA) Paired end reads: More precise mapping/alignment/variation calls Medium/Large indels (insertion/deletion) Structural variations Scaffolding in de novo assembly Scaffolding in de novo assembly Structural variations

13 Question What does it mean to have paired end reads? Discuss with neighbor for 2-3 mins, we discuss

14 Paired end reads x2 Illumina Paired End sequencing video

15 Exercise Data_basics_exercise.php