BRM5012 Chap.7 Single-cell Barcodes and UMIs
Single-cell Barcodes and UMIs
Single-cell RNA sequencing isolates viable cells, attaches a cell barcode to molecules from each cell and assigns a unique molecular identifier to each original RNA molecule. This chapter explains why those identifiers survive pooling, how reverse transcription and amplification create sequenceable cDNA, and how reads are reorganised into a cell by gene matrix without counting amplified copies as new molecules.
What this chapter covers
- 01
Cell isolation and dissociation
- 02
Cell barcodes
- 03
Unique molecular identifiers
- 04
Poly(T) priming
- 05
Reverse transcription
- 06
Amplification and pooling
- 07
Demultiplexing
- 08
nCount and nFeature
Worked example · free
Collapse repeated UMIs
- +1Use the shared cell barcode to place all reads in one cell.
- +1Collapse repeated u1 for Gene X and retain u1 and u2 as two original molecules.
- +1Count Gene Y u3 as one additional original molecule, giving nCount of three.
- +1Count Gene X and Gene Y as two detected genes, giving nFeature of two.
Key terms
- Cell barcode
- A tag shared by molecules from one cell that identifies their cellular origin after pooling.
- Unique molecular identifier
- A tag assigned to an original RNA molecule so amplified copies can be recognised and collapsed.
- Reverse transcription
- The conversion of RNA into complementary DNA that can be amplified and sequenced.
- Demultiplexing
- The computational separation of pooled reads by their retained sample or cell identifiers.
Single-cell Barcodes and UMIs FAQ
Why are both barcodes and UMIs needed?
The cell barcode preserves cell of origin after material is pooled. The UMI distinguishes original RNA molecules within that cell. Without the first, cells cannot be reconstructed; without the second, amplification can inflate abundance.
Can the same UMI occur in different cells?
Yes. UMIs are interpreted within a cell and gene context. Two reads with different cell barcodes are not duplicates merely because the short UMI sequence matches. Group by barcode before collapsing molecule identifiers.
What information is lost during dissociation?
Physical location and some tissue relationships are lost when cells are separated. Dissociation can also stress cells or underrepresent fragile populations. Preserve that limitation when interpreting the final matrix and consider spatial methods when location is central.
Assessment move
Draw one cell containing several RNA molecules, then attach the same barcode to every molecule and a different UMI to each original molecule. Follow those tags through lysis, reverse transcription, amplification, pooling and sequencing. Practise sorting synthetic reads in the correct hierarchy: cell barcode, gene, then UMI. Recompute nCount and nFeature from short examples and always name the unit of each total.
Review why poly(T) priming selects polyadenylated RNA and why amplification makes UMI handling necessary. For assessment explanations, link every molecular step to the information it preserves. Finish with a limitation such as lost spatial context, imperfect capture or cell-selection bias and explain how it affects the biological conclusion. Use coloured paper tags to model three cells before pooling.
Give every molecule from one cell the same barcode, assign UMIs independently, and copy selected molecules several times to represent amplification. Sort the pooled cards first by barcode, then gene, then UMI. Explain why matching UMIs in different cells are unrelated and why a UMI has duplicate meaning only after cell and feature are fixed. Build the corresponding gene-by-cell matrix by hand.
Sum each column for nCount and count its non-zero genes for nFeature, then compare these summaries with the original cards. Introduce one barcode error, one UMI error and one ambiguous gene assignment. Predict whether each mistake creates a false cell, inflates a molecule count or removes a count from the matrix. This makes pipeline correction rules easier to evaluate.
Trace what poly(T) priming selects and identify the RNA species or damaged molecules that may be underrepresented. Compare raw reads with counted molecules so amplification is never confused with biological abundance. For every zero, rehearse three possibilities: no expression, failed capture or insufficient observation. Draw the tissue before dissociation and mark the spatial relationship that the matrix cannot recover.
Then list cell types that handling might preferentially lose and states that dissociation might induce. In an assessment answer, describe the full reconstruction: viable-cell isolation, cell tagging, molecular tagging, cDNA production, amplification, pooling, sequencing, demultiplexing, feature assignment and UMI collapse.
Finish by naming which identifier preserves each layer of identity and one measurement boundary that survives all computation.
Working through Single-cell Barcodes and UMIs in BRM5012? Sia is AskSia’s AI Biology tutor — ask any BRM5012 Single-cell Barcodes and UMIs question and get a clear, step-by-step explanation grounded in how BRM5012 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
