Table 1.
Properties of methods for detecting hierarchies of TADs
| SuperTAD | OnTAD | deDoc | 3DNetMod | GMAP | IC-Finder | TADtree | Arrowhead | |
|---|---|---|---|---|---|---|---|---|
| Last updated | 2020 | 2019 | 2018 | 2018 | 2017 | 2017 | 2016 | 2014 |
| Journal | – | Genome Biology | Nature Communications | Nature | Nature Communications | Nucleic Acids Research | Bioinformatics | Cell |
| Input | Raw/normalized | Raw/normalized | Raw/normalized | Normalized | Normalized | Raw/normalized | Normalized | Raw/normalized |
| Model | Structural information theory, | Adaptive local minimum search, | Structure information theory, | Network modularity maximization, | Gaussian mixture model, | Hierarchical clustering, | Boundary index (BI), | Arrowhead transformation |
| dynamic programming | contact density signal maximization | greedy merging and combining | hierarchical spatial variance minimization | proportion test | local directionality index (DI) | dynamic programming, | ||
| weighted interval scheduling with | ||||||||
| multiplicities (WISM) | ||||||||
| Levels of TAD | Not limited | Not limited | 2 | Not limited | Controlled by parameter | Controlled by parameter | Not limited | Not limited |
| # Parameters | 0 | 2 | 0 | 5 | 4 | 1 | 6 | 1 |
| Reference | – | [34] | [33] | [32] | [31] | [30] | [29] | [28] |