Lecture 2

RNA seq reads are maped against the correspondinggenomic position

in a Perfect world, reads map only to exonic regions and split reads identify exon-intron boundaries

overlapping sequence reads with sufficient sequence similarity are collapsed into longer sequences (aka contigs)

the contigs serve as reconstruction of the original transcript

a set of reads that are related to one another by overlap of their sequences. all reads belon to one and only contig, and each contiguous contans at least one read

set of reads that locally overlap with each other

ordered and oriented - typically not overlapping - contigs seperated by gas approximtely known lentgh. Scaffolds ae typically fomred by identifying contig pairs that each contain one read of a `read pair`

De NoVo or Reference based

1. Reads

2. Read clean up

3. Assembly

   3.1 De novo:

   Trinity

   3.2 Reference based

   • alignment to reference genome

   • Transcript reconstruction

   1. Post assembly analysis

      • QC

      • Full length transcription analysis

      • Abundance estimation

      • DGE

      • Protein coding region annotation

      • Functional annotation

software can not distinguish between similar reads

if a kmer is found mak as used n the hash table

extended seed with the mew nucleotide

continue until no further extension possible

then start with unused Kmer as a new seed

decompose all reads into overlapping Kmers, inoring low-complexity kmers

integrate isoforms via k-1 overlaps, verify via “welds”

build de brujin graph