38
Canonical Grid model
Expression Groups and Read QC
Inspect synthetic expression groups, choose a reference distribution, and preserve named native analyses in the Sequence workspace.
- Scale
- Small
- Source
- biology-expression-review.grid
- Length
- 32 lines
- Collection
- Advanced language
- Level
- Intermediate
- Runtime
- Portable
- Version
- 1.2.0
MODEL "Expression Groups and Read QC"
DESCRIPTION "Inspect synthetic expression groups, choose a reference distribution, and preserve named native analyses in the Sequence workspace."
VERSION "1.2.0"
TAGS "biology", "bioinformatics", "expression", "quality control"
# Synthetic nonnegative expression scores, not raw RNA-seq counts.
# Observations 1–2 are controls; observations 3–4 are treated.
MatrixMarket = "%%MatrixMarket matrix coordinate real general\n4 3 8\n1 1 10\n1 2 4\n2 1 14\n2 2 6\n3 1 20\n3 2 8\n4 1 24\n4 2 10\n"
Expression = BIO_PARSE_MATRIX_MARKET(MatrixMarket)
# Formula exploration on a runtime supporting the explicit PCA contract.
input Components = 2
VALIDATE BETWEEN 1 AND 2 MESSAGE "Choose one or two components."
output CenteredScores = BIO_OMICS_EMBED_V2(Expression, Components, "pca")
output MatrixToPaste = MatrixMarket
output ReadsToPaste = "@read-one\nACGTACGT\n+\nIIIIIIII\n@read-two\nACGTACGT\n+\nIIII!!!!\n"
output StudyQuestion = "Does the reference distribution change the interpretation of a two-observation comparison?"
output Workflow = "In Sequence, open Paste scientific data. Paste MatrixToPaste as Matrix Market, choose observations in rows and general matrix, preview and import it. Choose Compare expression groups, name the analysis, set group A to rows 3,4 and group B to rows 1,2. Run Welch t, then repeat with a different name and the normal approximation. Paste ReadsToPaste as FASTQ, import it and choose Read quality control to inspect Q20/Q30 and cycle quality. Save the workbook archive to retain the native analyses."
output ReferenceAnswer = "For the first feature, the retained independent reference is Welch t p = 0.0715233091 with 2 degrees of freedom; the normal reference with corrected tail precision gives about 0.0004069520. These are different reference distributions, not interchangeable tests."
output Scenario = "Change Components from 2 to 1 to inspect a one-component formula result. In Sequence, compare Welch and normal references under separate names; choose the explicit PCA action for runtime-checked method support. Older runtimes refuse the versioned formula identity instead of ignoring its method. The read fixture has 2 reads, 16 bases and Q20/Q30 fractions 0.75."
# The Sequence workbench holds resident matrices and named native results.
Sequence_1!type = "sequence"
Sequence_1!config = <toml>
version = 1
kind = "sequence"
title = "Expression groups and read QC"
</toml>
END MODEL