Chromatin tracing

Multiplexed DNA FISH — ORCA, Hi-M, DNA-MERFISH, chromatin tracing in general — images one locus after another and gives each locus a position in 3-D. Its data arrive as a table of spots: one row per locus of one chromatin fibre, with x, y, z. That is a ChromData almost as it stands, so this chapter is mostly about getting the spots in, linking them into traces and filling the loci a trace missed; the analyses on the coordinates are common to all data and come after.

Step	API
read a 4DN FISH Omics Format table (FOF-CT), with its cell and RNA tables; write one	ChromData.from_fofct(core, cell_table=, rna_table=, out=), cd.to_fofct
read PyHiM traces (ECSV)	ChromData.from_pyhim_trace
read the tables of Bintu et al. 2018	uchrom.io.read_bintu_tracing
link raw detections into traces, with ploidy (the jie aligner)	uchrom.im.trace.align_spots
fill missing loci: linear, cubic spline, SnapFISH-IMPUTE	uchrom.im.impute.impute_coordinates(cd, method=...), uc.pp.impute

Tutorials

• Importing FOF-CT chromatin-tracing data
• Importing PyHiM chromatin-trace tables (ECSV)
• Spot-to-trace alignment with jie
• Imputing missing FISH coordinates

Guides

• The 4DN FISH Omics Format (FOF-CT)
• From detections to traces: the jie aligner

Then

Traces are coordinates, so everything that works on coordinates applies: structure calling from the distances between loci (loops, TADs, domains per allele, compartments), per-locus features (distance maps, contact frequencies, variances), the web browser. Imaging and Hi-C of the same cell type meet in GEM-FISH.

API: uchrom.im, uchrom.io.