Computational Systems Biology Methods and Protocols.7z

Table 1Computational tools for multistep analyses of scRNA-seq dataTool

Description

Requirements

Type

Interface

Operatingsystem

Programinglanguages References

Noise reductionscLVM

Uses a Gaussian Process Latent

variable model to estimate thecovariance matrix associatedwith latent factors. Residualsfrom a linear mixed modelwith the covariance termrepresent de-noisedexpression estimates

Requires genes associated with

the latent factor to beidentified a priori.Normalization factors areestimated using the mediannormalization method

Package

Command

line

Unix/

Linux

Python

[^79

]

f-scLVM

Use a factorial single-cell latent

variable model (f-scLVM) todissect and model single-celltranscriptome heterogeneity,thereby allowing to identifybiological drivers of cell-to-cell variability and modelconfounding factors

f-scLVM requires two input files,

a gene expression file and anannotation file. The geneexpression file is a text filecontaining the normalized,log-transformed geneexpression matrix, with everyrow corresponding to a cell.The annotation file is a text filewith every row containing thename of a gene set, followedby the gene identifiersannotated to that gene set

Package

Command

line

Unix/

Linux

Python

[^80

]

OEFinder

Uses orthogonal polynomial

regression to identify geneswhose expression is associatedwith position on the C1Fluidigm integrated fluidiccircuit (IFC)

Gene-specific P values are

provided to identify genesaffected by the artifact

Package

Command

line/GUI

Unix/

Linux,Mac OS,Windows

R[

81

]

(continued)

Applications of Single-Cell Sequencing for Multiomics 337