These are data for for Bein et al., 2020, NatComms (accepted): *Prior knowledge promotes hippocampal separation but cortical assimilation in the left inferior frontal gyrus* #### read here for explanation about the data Relevant code is abvailable on: https://github.com/odedbein/SEL_public I strongly advise to look there for details on analysis, if you want to reanalyse the data. ## Explanation about the data: #### In each subject folder you'll find: 1. raw_data: the raw epi scans files, this is really raw_data - no slice timing correction, no movement correction, and no normalization. Note that the scanner removes the first few dummy scans, so no need to cut any scans. In the raw_data folder, different folders for different scans: Similarity1/Similarity2 - the two scans for the pre learning similarity. Similarity3/Similarity4 - the two scans for the post-similarity. scans 1 and 3 are identical in the order of the stimuli, and scans 2 and 4 are identical (for pre-post differences, scans need to be identical, see paper). Pairs1-4: this is the associative learning task, it was divided to 4 scans (1-4). In the paper I report functional connectivity results, and in the supplementary, univariate analysis as well. 2. single_trial_Tmaps: I used single trials t-stats for the similarity analysis. If you want to use these - they are in this folder. these are from data that is slice timing and motion corrected, not smoothed, and in the subject native epi space. PRE_learning: t-stats from before the associative learning POST_learnig: t-stats from after the associative learning The data was modeled in spm from both scans in each pre/post, so the tstats numbers refer to tstats from both scans (that were averaged later in the analysis, see paper): for the PRE_learining: tstats 1-12: famous A faces tstats 13-24: novel A faces tstats 25-48: novel B faces that were paired with A faces. Each participant had their pairing, the pairing is provided in: ****** These numbers are from the first scan in the PRE_learning. tstats from the second scan are numbers 49-96, and they correspond exactly to the first scan, with a 48 difference. e.g., face1 in the first pre learning scan is face49 in the second pre learning scan. Same nubmers exactly in the POST_learning folder 3. mprage folder: this has the mprage T1-weighted mprage anatomical scan of each participant 4. behavior: this has the behavioral files from tasks. similarity: like the fMRI data - 1-2 are pre learning, 3-4 are post learning. A/B refers to two types of optseqs we had, and the order of A/B rotated across subjects. pairs: like the fMRI data 1-4 are the 4 scans. Here we had 4 optseq sequences, and the order rotated across participants. CR: that's the associative memory test data The numbers correspond to the fMRI data, and reflect the order in which participants performed the task: similarity1 (pre) similarity2 (pre) pairs1 pairs2 pairs3 pairs4 similarity3 (post) similarity4 (post) associative memory test (no imaging data, CR behavioral file) Notes on specific participants: If you run the analyses on fsl and don't need to register the runs to be aligned together, that's irrelevant. In SPM, since models combine runs, all runs are motion corrected to one space (I call it the meanimage). So I registered specifically for these subjects: 110715YB – did similarity1 and 2, left to the loo, went out of the scanner, and came back and continued. When I analysed the data, similarity1 and 2 were coregistered to the other sessions. 250715EB – left to the loo between pairs4 and similarity 3, I coregistered similarity 3,4 to the pairs and similairty1,2 sessions. Subjects 200615TF and 230615ZD were scanned bottom-up interleaved due to a scanner mistake (all other subjects were scanned top-down interleaved). If you do slice timing correciton, correct accordingly.