Fluorescence in situ hybridization (FISH) is an established technique by which fluorescently labeled DNA strands are hybridized to target nucleotide molecules and imaged using a fluorescence microscope. While locus-specific FISH probes only generate data on the loci they target, chromosome paints enable the study of whole chromosomes. Further, a multi-color set of paints can be used to target the entire genome at once. This is the case for assays such as spectral karyotyping (SKY)1, combinatorial multi-fluor FISH (mFISH)2, and directional Genomic Hybridization (dGH) SCREEN3,4. These are powerful techniques, especially because they offer single-cell, whole-genome data about the entire chromosome complement via direct visualization. However, one thing these assays do not provide is the precise breakpoints for any rearrangements detected, only that a structural variant has affected a certain arm of a chromosome. FISH-based banding assays were developed with the aim of addressing this need. Here, we will contrast two such assays: multicolor-banding (mBAND or MCB)5 and dGH DSCVR6.

It is worth asking where FISH-based techniques stand generally, in the context of powerful sequencing technologies. While the most extreme examples of long-read sequencing have achieved read lengths of over a megabase, the norm is in the range of tens to hundreds of kilobases. Aside from the technology being costly, long-read sequencing also forces users to choose between throughput and accuracy. Long-read sequencing can have a relatively high error-rate. This can be reduced by increasing read depth, but that raises time and cost. Additionally, data generated by all sequencing technologies requires an alignment step, producing results which vary slightly from one alignment algorithm to the next. By contrast, FISH based assays involve looking at single cells directly. This yields high-confidence data and also means the context in which each variant is detected is known. Contextual information includes whether a detected abnormality is heterozygous or homozygous, what other abnormalities occur with it, how frequently, etc. This information is either out of reach or carries varying levels of uncertainty when sequencing DNA pooled from multiple cells. While single-cell sequencing can partly address these data gaps, it increases cost and time. Interpreting single-cell data can also be difficult when characteristics of the individual sequencing reactions are not constant, such as with droplet size variability. FISH-based assays can help address these data gaps and uncertainties.

mBAND is a relatively straight forward assay. While a paint probe labels an entire chromosome in a solid color, mBAND subdivides that probe set. The chromosome is hybridized with a series of probe sets targeting sections of the chromosome from top to bottom. Each probe set is conjugated with a different blend of fluorochromes in a pre-defined sequence. The result is a known banding pattern in which structural variants are revealed by segments displaying rearrangements in the expected false-color sequence. While exact breakpoints still can’t be known, changes as small as 1-2 Mb can be detected and roughly localized. For context, G-banding, also based on changes to a chromosome’s expected banding pattern, detects rearrangements as small as 5-10 Mb.

Example of mBAND chromosomes. (Source: Chudoba, et al. 1999.)

dGH DSCVR, as with all dGH assays, includes some changes at the cell culture stage. While cells are still dividing, analog nucleotides are added to the culture medium. These are incorporated into newly synthesized strands of DNA. When the harvested cells are later subjected to UV irradiation, these daughter strands are more vulnerable to nicking compared with the parent strands. The nicks allow their removal through enzymatic degradation, leaving chromosomes composed of only parent strands. Banded paint probes similar to those of mBAND are then designed, but only sequences on one of the two parent strands are targeted by the probe set. The outcome is a set of banded chromosomes in which only one chromatid is painted.

Example of dGH DSCVR karyogram. (Source: KromaTiD, Biomarker Discovery & Target Validation.)

dGH DSCVR has a critical advantage over mBAND. Aside from being higher-resolution, able to detect changes as small as 5-10 kilobases, it can spot small inversions. Inversions are a challenge for many platforms. They do not necessarily change the copy number, and the link between the breakpoint on one end of the rearrangement and on the other end may not be obvious. However, inversions in dGH chromosomes cause fluorescence to jump from the painted chromatid to the unpainted one. This makes even very small events easy to spot, as can be seen in the example image: at the top of both copies of chromosome 8, some of the red fluorescence has relocated to the chromatid which should appear solidly blue. dGH DSCVR will detect any other type of rearrangement as well: translocations, deletions, insertions, etc. However, it is the data on small inversions, which can be especially challenging to obtain.

As newer and more powerful genomic techniques are progressively developed, it is important to know each method’s limitations, where the data gaps are. Every technique has some blind spot, some disadvantage, some inefficiency or impracticality. Merging data generated by disparate, orthogonal technologies gradually produces a more complete picture of what’s happening in a cell, in a cell population, or in a sample set. This strengthens data confidence and can lead to greater patient safety down the development pipeline.



  1. Imataka, G., & Arisaka, O. (2012). Chromosome Analysis Using Spectral Karyotyping (SKY). Cell Biochemistry and Biophysics, 62(1), 13-17. https://doi.org/10.1007/s12013-011-9285-2
  2. Speicher, M. R., Ballard, S. G., & Ward, D. C. (1996). Karyotyping Human Chromosomes By Combinatorial Multi-Fluor FISH. Nature Genetics, 12(4), 368-375.
  3. McKenna, M. J., Robinson, E., Goodwin, E. H., Cornforth, M. N., & Bailey, S. M. (2017). Telomeres and NextGen CO-FISH: Directional Genomic Hybridization (Telo-dGH™). Methods in Molecular Biology, 1587, 103-112. https://doi.org/10.1007/978-1-4939-6892-3_10
  4. KromaTiD, Inc. (2024). dGH SCREEN™ Assay Services. Retrieved from https://kromatid.com/high-resolution-structural-variant-data/
  5. Chudoba, I., Plesch, A., Lörch, T., Lemke, J., Claussen, U., & Sengera, G. (1999). High Resolution Multicolor-Banding: A New Technique for Refined FISH Analysis of Human Chromosomes. Cytogenet Cell Genet, 84, 156-160.
  6. KromaTiD, Inc. (2024). Biomarker Discovery & Target Validation. Retrieved from https://kromatid.com/biomarker-discovery-target-validation/
  7. Liehr, T (Ed.) (2017). Fluorescence In Situ Hybridization (FISH) Application Guide (2nd ed., p. 241).