Data associated with the submitted manuscript "Resolving Functional States in Cryo-EM Datasets with JANAS"

Here we include datasets associated with the application of JANAS (Joint ANAlysis of Stacks), a cryo-EM data processing pipeline for particle selection and 3D class reassignment based on the Structural Cross-correlation Index (SCI). The deposited data include: EMPIAR-10308 (5-HT1BR–Go receptor complex), used to assess SCI-based global particle selection and a focused analysis of the ligand-binding pocket; EMPIAR-10667 (46Q HTT–HAP40 complex), used to evaluate class reassignment on the deposited stack and reproducibility on an independently repicked stack; a newly acquired O. iheyensis group II intron dataset; the OLE RNA homodimer (EMPIAR-12707), used as an RNA-only benchmark; and an SCI_test archive providing the source data for the SCI characterisation and method-comparison experiments. Raw particle stacks from EMPIAR-10308, EMPIAR-10667 and EMPIAR-12707 are already publicly deposited and thus not included here. For the group II intron dataset, particle stacks for J317 and J246 are provided. Datasets also include the refined 3D maps, atomic models and related metadata. EMPIAR_10308_particle_selection.zip This archive contains the SCI-based particle selection results applied to EMPIAR-10308 (5-HT1BR–Go receptor complex bound to donitriptan). It also includes half-maps, ModelAngelo automatic models obtained from the unprocessed maps, and the *_stackNotationEMPIAR.star files for particle stacks corresponding to the subset of particles ranked and selected. Directories include deposited_data, corresponding to the original EMPIAR-10308 deposit; selected_244565_particles, with the selected subset; and selected_244565_particles_refined, with the angular-refined maps reconstructed from the selected particles. EMPIAR_10308_ligand.zip This archive includes focused classification and class-reassignment data for the ligand-binding region of the 5-HT1BR–Go receptor complex from EMPIAR-10308. Files include particle stacks, class assignments and refined maps corresponding to the five final classes identified in Fig. S4 of the manuscript, in support of a comparative evaluation of conformational states within the binding pocket, and to the two selected classes carried forward for ligand modelling: the Ser334-proximal pose (16,736 particles) and the linker-reoriented pose (15,349 particles). The archive contains refined maps (half-maps and post-processed maps), local-resolution estimations, masks, the corresponding .star files of particles contributing to each class, the cryoSPARC reference reconstructions used as input, the signal-subtracted reconstructions, and the two new atomic models of the ligand-binding pocket (Ser334.pdb, Linker-reoriented.pdb) together with the deposited reference model (PDB 6G79). EMPIAR_10667_final_classes_selected.zip and EMPIAR_10667_3dFSC_analysis.zip These archives contain the class reassignment and 3D refinement results for EMPIAR-10667 (46Q HTT–HAP40 complex). Particular focus is given to flexible regions such as the N-terminal BΦ motif, and domain-resolved refinements are available for further comparative analysis. Included files are final refined maps for each selected class, ModelAngelo traces for each selected class, .star files of selected particles, local resolution maps, directional 3D FSC volumes and correlation plots, and atomic models corresponding to each class, including mask and map-fitting metadata. EMPIAR_10667_repick.zip This archive contains the SCI-seeded repicking and class-reassignment results for EMPIAR-10667, testing the reproducibility of the BΦ-motif conformations through an independent upstream coordinate-generation route (Fig. S7 of the manuscript). A small SCI-scored, orientation-balanced subset was used to seed cryoSPARC template matching and Topaz repicking from the original micrographs, yielding a stack of 458,445 particles with reduced preferred orientation, which was then processed by non-uniform refinement, 3D class reassignment and per-class selection. Included files are the seeding subset, the topaz-repicked metadata and associated cryoSPARC reconstruction, the masks used for reassignment, the per-class reassignment outputs, and the per-class selected reconstructions with half-maps, local-resolution maps and atomic models, together with the reassignment and selection scripts. groupII_intron.zip This dataset includes newly acquired cryo-EM data for the Oceanobacillus iheyensis group II intron, with the aim of facilitating detailed exploration of conformational heterogeneity in the catalytic core and peripheral regions. The archive contains particle stacks for J317 and J246. Further, it contains the seven refined 3D classes and associated data: metadata .star files of particles assigned to each class, full and half maps for all classes, masks, local resolution maps and FSC plots, validation reports for deposited maps on the PDB, a data processing summary, and comparative resolution data. EMPIAR_12707.zip This archive contains the SCI-based particle-selection results for the OLE RNA homodimer (EMPIAR-12707; deposited map EMD-48163; deposited model PDB 9MCW), a 373 kDa RNA-only assembly used as an additional benchmark with a stable central scaffold and local-resolution variation in peripheral regions associated with the proposed OapA/OapC binding architecture (Fig. 7 of the manuscript). Three iterative rounds of particle selection with adaptive masking and angular refinement between rounds retained 64,481 of the initial 87,716 particles. Included files are: the particle metadata (.star files) for the full deposited stack and for the selected subset; the reconstruction and half-maps of the selected subset; reconstructions of the major 3D classes obtained from the cumulatively excluded particles (supporting Fig. S12); the adaptive mask and ChimeraX session; and the atomic model refined against the selected reconstruction starting from PDB 9MCW. SCI_test.zip This archive provides the source data and supporting files for the SCI characterisation and method-comparison experiments. It supports the SCI vs LoG sensitivity test on a controlled image with known geometric perturbations (Fig. S1B), the defocus-resampling control on EMPIAR-10308 (Fig. S2), and the rank-correlation analysis between SCI and CC, NMI, SSIM and MVPD on EMPIAR-10308 and EMPIAR-10667 (Supplementary Results 4.1). Included files are: the rotated reference images and processing script for the LoG sensitivity test; the per-method 10,000-particle subsets used for the defocus-resampling control; the Euler-balanced normalised rank .star files for each metric on both datasets, together with the reconstructions and masks used for ranking; and the scripts for computing the per-dataset Spearman ρ and Kendall τ rank-correlation values reported in the manuscript.