experimental planning

10X platform:

• UMIs integrated

• higher throughput leading to way more cells sequenced

• higher dropout rate, noisier data

• low coverage

• only 3’ data

• cheaper

SmartSeq:

• sorting required (e.g. FACS)

• higher sequencing depth

• more sensititive gene detection

• good if few cells need to be deeply characterized (precious samples)

• full length transcripts

  
  

• Oil suspension to get single-cells including beads (emulsion)

• Beads are added that contain R1, 10X barcode and polydTVN sequences

• 
  

The capture rate is 1 - the fraction of cells that that are in droplets without any beads.

The split rate is the fraction of droplets with exactly one cell that have more than one bead.

• The doublet rate is the fraction of droplets with 1 bead that have more than one cell.

Fraction of transcripts detected in every cell

• I1.fastq - sample index (more or less experimental/sample id, 10X barcodes are re-usable and just enough for roughly 700-800k cells)

• R1.fastq files - cell barcode + UMI

• R2.fastq files - transcript information

• Read Trimming:

• STAR Alignment:

  • MAPQ filtering

  • filtering on uniquely mapped reads

• 10x barcode correction:

  • Count the observed frequency of every barcode on the whitelist in the dataset.

  • For every observed barcode in the dataset that is not on the whitelist and is at most one Hamming distance away from the whitelist sequences:

    • Compute the posterior probability that the observed barcode did originate from the whitelist barcode but has a sequencing error at the differing base (by base quality score).

    • Replace the observed barcode with the whitelist barcode that has the highest posterior probability (>0.975).

• UMI counting:

  • If two groups of reads have the same barcode and gene, but their UMIs differ by a single base (i.e., are one Hamming distance apart), then one of the UMIs was likely introduced by a substitution error in sequencing. In this case, the UMI of the less-supported read group is corrected to the UMI with higher support.