Improved accuracy assessment for 3D genome reconstructions.

Area of Science:

• Genomics
• Molecular Biology
• Biophysics

Background:

• Three-dimensional (3D) genome organization influences cellular functions like transcription and can drive oncogenesis.
• Chromatin conformation capture assays, especially Hi-C, have advanced understanding of chromatin architecture.
• Assessing the accuracy of 3D genome reconstructions derived from proximity data remains a significant challenge.

Purpose of the Study:

• To develop and validate novel methodologies for assessing the accuracy of 3D genome reconstructions.
• To address the lack of gold standards for evaluating the fidelity of reconstructed genome structures.
• To investigate chromosome-specific accuracy in 3D genome reconstructions.

Main Methods:

• Utilized multiplex FISH with an increased number of probes to enable structure-level evaluation via mean-squared deviations (MSD).
• Employed large numbers of coordinate-system aligned replicates for referent distribution in MSD statistics.
• Applied genome architecture mapping, leveraging cryosection planarity for accuracy assessment with resampling-based referents.

Main Results:

• Demonstrated that Hi-C based 3D genome reconstructions for IMR90 cells exhibit varying accuracy across different chromosomes.
• Showcased varying reconstruction accuracies by chromosome in mouse embryonic stem cells using genome architecture mapping.
• Validated the utility of multiplex FISH and genome architecture mapping for assessing 3D reconstruction fidelity.

Conclusions:

• Developed robust methods for assessing 3D genome reconstruction accuracy by integrating multiplex FISH and genome architecture mapping.
• These novel approaches provide crucial tools for evaluating the reliability of 3D genome models.
• The findings highlight the need for rigorous accuracy assessment in 3D genome organization studies.