OSCR

Using connectome-based predictive models to reveal the systems standardized tests and clinical symptoms are reflecting.

Code ↔ Paper

6 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 6 matches
  1. [1] § Methods › Cognitive constructs & a priori network definitions ↔ Testing_Set227_KRR.m, lines 60–120 · score 0.86 · Shen268 atlas, declarative memory, network definition, working memory, construct nodes, cognitive control
  2. [2] § Results ↔ Testing_Set227_KRR.m, lines 60–120 · score 0.84 · NeuroSynth, Shen268 atlas, declarative memory, working memory, connectivity matrices, cognitive control
  3. [3] § Methods › Cognitive constructs & a priori network definitions ↔ Testing_Set227_KRR_combined_scores.m, lines 42–109 · score 0.76 · Shen268 atlas, declarative memory, network definition, construct nodes, hippocampus, ROI
  4. [4] § Results ↔ Testing_Set227_KRR.m, lines 1–58 · score 0.61 · kernel ridge regression, predicted behavioral, connectivity matrices, behavioral scores, edges, modeling
  5. [5] § Methods › Model ↔ Testing_Set227_KRR.m, lines 1–58 · score 0.59 · kernel ridge regression, prediction errors, connectivity matrices, covariate, edges, models
  6. [6] § Results › Testing the tests can be performed with any network of interest ↔ Testing_Set227_KRR_combined_scores.m, lines 42–109 · score 0.55 · Yeo networks, network definitions, language network, DMN, atlas, brain

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

MATLAB · 493 lines · 23 KB · no license · 4 matches

  1. % This is an example to show how to run the kernel ridge regression model
  2. % to predict behavioral scores based on the connectivity matrices.
  3. % We provided a subset of data that was collected and processed at Yale MRRC.
  4. % The provided dataset includes connectivity matrices, behavioral scores, and
  5. % covariates from all subjects.
  7. clear
  8. clc
  10. % % define your data path
  11. % data_path = '';
  12. % % define your path to save the results
  13. % save_path = '';
  14. % % define your path to save the prediction errors
  15. % error_path = '';
  18. % load the matrix after the regression
  19. % Each edge in the matrix is the residue after regressing out the covariates
  20. load([data_path, 'trans_data_set227_mat_cov_regressed'], 'res_mat');
  22. % load the original scores
  23. load([data_path, 'trans_data_all_scores_set227'],'all_score');
  24. % 146 = 124 + 16 + 6; other sub-scores + cognitive_scores + clinical scores
  25. % the first 124 columns are sub-scores
  26. % column 125-140 are the 16 cognitive scores
  27. % column 141-146 are the 6 clinical scores
  29. % load the regressed scores (for evaluation purpose)
  30. % the covariates are regressed from the behavioral scores
  31. load([data_path, 'trans_data_set227_allscores_cov_regressed'], 'all_res');
  33. % load the covariates
  34. load([data_path, 'trans_data_combined_update_cov_diag_set227'], 'cov_combined');
  37. % choose the set of behavioral scores to predict
  38. nan_flg = 1;
  39. if( nan_flg ==1)
  40. % cognitive scores (16)
  41. datamat_orig = all_score(:, 125:140);
  42. datamat_res = all_res(:, 125:140);
  43. elseif( nan_flg==2)
  44. % all rest of the sub scores (124)
  45. datamat_orig = all_score(:, 1:124);
  46. datamat_res = all_res(:, 1:124);
  47. elseif( nan_flg ==3)
  48. % clinical scores (6)
  49. datamat_orig = all_score(:, 141:146);
  50. datamat_res = all_res(:, 141: 146);
  51. elseif( nan_flg ==4)
  52. % load the newly computed brief composite score
  53. % we selected a subset of the brief subscore and create a new composite score
  54. load([data_path, 'trans_data_set227_brief_new_composite'], 'new_brief_orig', 'new_brief_res', 'new_name');
  56. datamat_orig = new_brief_orig;
  57. datamat_res = new_brief_res;
  58. end
  60. % find the number of behavioral scores to predict
  61. no_beh = size( datamat_res, 2);
  63. % matrices after regression
  64. multi_run_all_mats = res_mat;
  65. % the connectivity matrices are organized as a 4D matrix of no_of_nodes * no_of_nodes * no_runs * no_subjects
  66. no_runs = size(multi_run_all_mats, 3); % each subject has 8 runs
  67. no_sub = size(multi_run_all_mats, 4);
  68. no_nodes = size(multi_run_all_mats,1);
  69. % the atlas we used for creating the connectivity matrices has 268 nodes.
  71. % find the indices of the edges in the upper triangle of a matrix
  72. % the connectivity matrix is symmetric
  73. aa = ones( no_nodes, no_nodes);
  74. aa_upp = triu(aa,1);
  75. upp_id = find( aa_upp>0);
  76. upp_len = length(upp_id);
  78. % To select edges from pre-defined networks/constructs
  79. % The prediction is based on an individual network/construct
  80. % yeo_flg = 1; % the yeo network (7 networks)
  81. % yeo_flg = 3; % use DMN and Ev language
  82. % yeo_flg = 2; % use the whole brain
  83. yeo_flg = 0; % use the constructs (6 constructs defined based on NeuroSynth: attention, perception, declarative memory, language, cognitive control and working memory)
  84. % yeo_flg = 4; % combine all 6 constructs;
  86. if( yeo_flg==0)
  87. % load the construct nodes
  88. load([data_path, 'construct6_network268_node_list'], 'check_roi');
  89. % 'check_roi' was created based on NeuroSynth results
  90. % it is a labeling matrix where each column represents one construct
  91. % and each row represents one ROI
  92. % the order of the columns: attention, perception, declarative memory, language, cognitive control, working memory
  94. % replce the declarative memory construct with nodes from the
  95. % hippocampus from the Shen268 atlas
  96. manual_dec = 1;
  97. % including bilateral hippocampus nodes
  98. if(manual_dec==1)
  99. hip_id = [93:97 230:234];
  100. check_roi(:, 3) = 0;
  101. check_roi(hip_id, 3) = 1;
  102. end
  103. no_net = 6;
  104. elseif( yeo_flg ==1)
  105. % load the Yeo's 7 network definition
  107. yeomap = dlmread([data_path, 'Parc268toYeo7netlabel']);
  109. no_net = size( yeomap, 2);
  110. check_roi = yeomap;
  111. elseif( yeo_flg ==2)
  112. % use the whole brain;
  113. elseif( yeo_flg ==3)
  114. % use the dmn and Ev's language network
  115. file_name = 'Shen268_10network';
  116. % use the dmn network definition from the in-house developed resting state functional brain network
  117. n_label = dlmread([data_path, file_name], '\t', 0, 0);
  118. dmn_nodes = find( n_label(:,2)==3);
  119. check_roi = zeros(no_nodes, 2);
  120. check_roi( n_label(dmn_nodes,1), 1) = 1;
  122. % use the language network:
  123. % Lipkin, Ben (2022). LanA Dataset. figshare. Dataset. https://doi.org/10.6084/m9.figshare.20425209.v1
  124. load([data_path,'Ev_lan_nodes_12'], 'lan_node');
  125. check_roi( lan_node, 2) =1;
  127. no_net = 2;
  128. elseif( yeo_flg==4)
  129. % load the construct nodes and combine all of them
  130. load([data_path, 'construct6_network268_node_list'], 'check_roi');
  131. manual_dec = 1; %
  132. if(manual_dec==1)
  133. hip_id = [93:97 230:234];
  134. check_roi(:, 3) = 0;
  135. check_roi(hip_id, 3) = 1;
  136. end
  138. sum_check = sum( check_roi,2);
  139. check_roi = (sum_check>0);
  140. no_net = 1;
  141. end
  143. % filling the lower triangle of the conn matrix
  144. for run_idx = 1: no_runs
  145. for sub_id = 1: no_sub
  146. cur = squeeze(multi_run_all_mats(:,:,run_idx, sub_id));
  147. cur = cur+transpose(cur);
  148. multi_run_all_mats(:,:, run_idx, sub_id) = cur;
  149. end
  150. end
  152. % lambda is the hyper-parameter of the kernle ridge regression
  153. lambda = [ 0 0.00001 0.0001 0.001 0.004 0.007 0.01 0.04 0.07 0.1 0.4 0.7 1 1.5 2 2.5 3 3.5 4 5 10 15 20];
  154. lambda_no = length(lambda);
  156. no_iter = 1; % number of iteration for the k-fold cross validation
  157. scale_flg =0; % scale of the kernel
  158. fold = 10; % k-fold
  159. bin_size = round( no_sub/fold); % size of each fold
  161. error_flg = 0; % whether or not to save the prediction error
  163. for ll = 1: lambda_no % loop through all lambda values
  165. disp(['lambda = ', num2str(lambda(ll))]);
  166. cur_lambda = lambda(ll);
  168. % create placeholders for prediction performance
  169. if( yeo_flg <2)
  170. all_perf_r = zeros(no_iter, no_beh, no_net);
  171. all_perf_p = zeros(no_iter, no_beh, no_net);
  172. elseif( yeo_flg==2)
  173. all_perf_r = zeros(no_iter, no_beh);
  174. all_perf_p = zeros(no_iter, no_beh);
  175. elseif( yeo_flg==3)
  176. all_perf_r = zeros(no_iter, no_beh, no_net);
  177. all_perf_p = zeros(no_iter, no_beh, no_net);
  178. end
  180. % start the prediction
  181. for iter = 1: no_iter
  183. cur_order= randperm(no_sub); % shuffle the subject order
  184. cur_cov = cov_combined(cur_order, :); % shuffle the covariates accordingly
  186. % predict a single behavioral score
  187. for behav_idx = 1: no_beh;
  189. % select the behavioral score to predict
  190. all_behav_res = datamat_res(:, behav_idx);
  191. all_behav_orig = datamat_orig(:, behav_idx);
  193. % first shuffle the behavioral score
  194. all_behav_res = all_behav_res(cur_order);
  195. all_behav_orig = all_behav_orig(cur_order);
  197. % then remove subjects with NaN scores
  198. inner_va_id = find( isnan(all_behav_orig)==0);
  199. cur_no_sub = length(inner_va_id);
  201. % all with valid scores
  202. cur_behav_orig = all_behav_orig(inner_va_id);
  203. cur_behav_res = all_behav_res(inner_va_id);
  205. cur_cov_va = cur_cov(inner_va_id, :); % the 5 covariates
  206. va_order = cur_order(inner_va_id);
  208. % placeholders to save the predicted values
  209. if( yeo_flg <2 || yeo_flg>2)
  210. behav_pred = zeros(cur_no_sub,no_net);
  211. else
  212. behav_pred = zeros(cur_no_sub, 1);
  213. end
  215. % using all runs
  216. all_mats = multi_run_all_mats(:,:,:, cur_order);
  217. cur_mat = all_mats(:,:, :, inner_va_id);
  219. % start the cross validation loop
  220. for leftout = 1:fold;
  222. left_sub = (leftout-1)*bin_size+1: min(cur_no_sub, leftout*bin_size);
  224. % leave out subjects from matrices and behavior
  226. train_mats = cur_mat;
  227. train_mats(:,:,:, left_sub) = [];
  228. % number of training subjects
  229. c_sub_no = size(train_mats, 4);
  231. % subject id included in the training set
  232. cur_train_sub = va_order;
  233. cur_train_sub(left_sub) = [];
  235. % regress the covariates within the training set
  236. train_cov = cur_cov_va;
  237. train_cov(left_sub, :) = [];
  239. train_behav_orig = cur_behav_orig;
  240. train_behav_orig(left_sub) = [];
  242. [b, bint, train_behav_res] = regress( train_behav_orig, [train_cov, ones(c_sub_no, 1)]);
  244. % normalize the score after regression
  245. train_mu = mean(train_behav_res);
  246. train_std = std(train_behav_res);
  247. train_norm = (train_behav_res-train_mu)/train_std;
  249. % matrices of the test subjects
  250. test_mat = cur_mat(:,:,:, left_sub);
  252. left_no = length( left_sub);
  254. if( yeo_flg <2 || yeo_flg>2)
  255. for con = 1: no_net;
  256. % find the nodes for the selected network/construct
  257. if( yeo_flg ==0)
  258. cur_con_nodes = find( check_roi(:, con)>0);
  259. elseif( yeo_flg ==1)
  260. cur_con_nodes = find( check_roi(:, con)==1);
  261. elseif( yeo_flg== 3)
  262. cur_con_nodes = find( check_roi(:, con)==1);
  263. elseif( yeo_flg==4)
  264. cur_con_nodes = find( check_roi(:, con)==1);
  265. end
  266. cur_no = length( cur_con_nodes);
  268. % find the edges between the selected nodes and the rest of the brain
  269. all_nodes = ones(1, no_nodes);
  270. all_nodes(cur_con_nodes) = 0;
  271. all_d_nodes = find( all_nodes==1);
  272. aa = ones(no_nodes, no_nodes);
  273. aa_upp = triu(aa, 1);
  274. aa_upp(all_d_nodes, all_d_nodes) = 0;
  275. upp_id = find( aa_upp);
  277. cur_con_mat = [];
  278. cur_con_test = [];
  280. % running the prediction from each individual runs
  281. for run_idx = 1:no_runs
  282. cur_train_mat = reshape(squeeze(train_mats(:,:,run_idx,:)), no_nodes*no_nodes, c_sub_no);
  283. cur_con_mat = [cur_con_mat; cur_train_mat(upp_id,:)];
  285. cur_test_mat = reshape( squeeze(test_mat(:,:,run_idx,:)), no_nodes*no_nodes, left_no);
  286. cur_con_test = [cur_con_test; cur_test_mat(upp_id,:)];
  287. end
  289. % scale of the Gaussian kernel
  290. if( scale_flg==0)
  291. scale = 0.2;
  292. else
  293. scale = 1;
  294. end
  296. % running the kernel ridge regression
  297. [pred, estimate] = kernel_prediction(cur_con_mat, cur_con_test, train_norm, cur_lambda, 'Gaussian', scale);
  298. behav_pred(left_sub, con) = pred;
  299. end
  300. elseif( yeo_flg ==2)
  301. cur_con_mat =[];
  302. cur_con_test = [];
  304. % concatenate the edges from all 8 runs
  305. for run_idx = 1:no_runs
  306. cur_train_mat = reshape(squeeze(train_mats(:,:,run_idx,:)), no_nodes*no_nodes, c_sub_no);
  307. cur_con_mat = [cur_con_mat; cur_train_mat(upp_id,:)];
  309. cur_test_mat = reshape( squeeze(test_mat(:,:,run_idx,:)), no_nodes*no_nodes, left_no);
  310. cur_con_test = [cur_con_test; cur_test_mat(upp_id,:)];
  311. end
  313. if( scale_flg==0)
  314. scale = 0.2;
  315. else
  316. scale = 1;
  317. end
  319. [pred, estimate] = kernel_prediction(cur_con_mat, cur_con_test, train_norm, cur_lambda, 'Gaussian', scale);
  320. behav_pred(left_sub) = pred;
  321. end
  323. end
  325. if(yeo_flg<2 || yeo_flg>2)
  326. pred_oo = zeros(no_sub, no_net); % prediction before the shuffle
  327. error_oo = zeros(no_sub, no_net); % error before the shuffle
  328. for con= 1:no_net;
  329. % compute the correlation between the predicted scores and the observed scores (after regression)
  330. [R, P] = corr( squeeze(behav_pred(:, con)), cur_behav_res);
  331. all_perf_r(iter, behav_idx, con) = R;
  332. all_perf_p(iter , behav_idx, con) = P;
  334. % compute the prediction error
  335. pred_ss = zeros(1, no_sub); % prediction after the shuffle
  336. pred_ss(inner_va_id) = behav_pred(:,con);
  337. % prediction error in the original order of subjects
  338. pred_oo(cur_order, con) = pred_ss;
  340. error_ss = zeros(1, no_sub);
  341. cc_mean = mean(cur_behav_res);
  342. cur_behav_mean = cur_behav_res-cc_mean;
  343. cc_norm = norm(cur_behav_mean); % normalized the scores (after regression)
  344. error_ss(inner_va_id) = (behav_pred(:, con) - cur_behav_mean/cc_norm); % signed error
  345. error_oo(cur_order, con) = error_ss;
  347. clear cc_norm cc_mean;
  348. end
  349. else
  350. pred_oo = zeros(no_sub, 1);
  351. error_oo = zeros(no_sub, 1);
  353. % compute the correlation between the predicted scores and the observed scores (after regression)
  354. [R, P] = corr( behav_pred, cur_behav_res);
  355. all_perf_r(iter, behav_idx) = R;
  356. all_perf_p(iter, behav_idx) = P;
  358. % compute the prediction error
  359. pred_ss = zeros( no_sub, 1); % prediction after the shuffle
  360. pred_ss(inner_va_id) = behav_pred;
  361. % in the original order
  362. pred_oo(cur_order) = pred_ss;
  364. error_ss = zeros(no_sub, 1);
  365. cc_mean = mean(cur_behav_res);
  366. cur_behav_mean = cur_behav_res-cc_mean;
  367. cc_norm = norm(cur_behav_mean);
  368. error_ss(inner_va_id) = (behav_pred - cur_behav_mean/cc_norm); % signed error
  369. error_oo(cur_order) = error_ss;
  371. clear cc_norm cc_mean;
  372. end
  374. % save the prediction errors
  375. if( error_flg ==1)
  376. error_path = '/data22/mri_group/xilin_data/alex/trans_data/set227/errors/';
  377. if( yeo_flg==0)
  378. if( nan_flg ==1)
  379. save([error_path,'trans_data_set227_KRR_error_singlecon_cogbehav', num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  380. elseif( nan_flg ==2)
  381. save([error_path,'trans_data_set227_KRR_error_singlecon_subscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  382. elseif( nan_flg==3)
  383. save([error_path,'trans_data_set227_KRR_error_singlecon_clicscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  384. end
  385. elseif( yeo_flg ==1)
  386. if( nan_flg ==1)
  387. save([error_path,'trans_data_set227_KRR_error_yeonet_cogbehav', num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  388. elseif( nan_flg ==2)
  389. save([error_path,'trans_data_set227_KRR_error_yeonet_subscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  390. elseif( nan_flg==3)
  391. save([error_path,'trans_data_set227_KRR_error_yeonet_clicscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  392. end
  393. elseif( yeo_flg==2)
  394. if( nan_flg ==1)
  395. save([error_path,'trans_data_set227_KRR_error_wholebrain_cogbehav', num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  396. elseif( nan_flg ==2)
  397. save([error_path,'trans_data_set227_KRR_error_wholebrain_subscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  398. elseif( nan_flg==3)
  399. save([error_path,'trans_data_set227_KRR_error_wholebrain_clicscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  400. end
  401. elseif( yeo_flg==3)
  402. if( nan_flg ==1)
  403. save([error_path,'trans_data_set227_KRR_error_dmnEv_cogbehav', num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  404. elseif( nan_flg ==2)
  405. save([error_path,'trans_data_set227_KRR_error_dmnEv_subscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  406. elseif( nan_flg==3)
  407. save([error_path,'trans_data_set227_KRR_error_dmnEv_clicscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  408. end
  409. elseif( yeo_flg ==4)
  410. if( nan_flg ==1)
  411. save([error_path,'trans_data_set227_KRR_error_all6con_cogbehav', num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  412. elseif( nan_flg ==2)
  413. save([error_path,'trans_data_set227_KRR_error_all6con_subscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  414. elseif( nan_flg==3)
  415. save([error_path,'trans_data_set227_KRR_error_all6con_clicscores',num2str(behav_idx), '_iter', num2str(iter)], 'error_oo', 'pred_oo', 'cur_order');
  416. end
  417. end
  418. end
  419. end
  420. end
  421. % save the prediction performance
  422. if( yeo_flg ==0)
  423. if( scale_flg ==0)
  424. if( nan_flg==1)
  425. save([save_path, 'trans_data_singlecon_KRR_10fold_normalized_allruns_set227_cogbehav_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  426. elseif( nan_flg==2)
  427. save([save_path, 'trans_data_singlecon_KRR_10fold_normalized_allruns_set227_subscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  428. elseif( nan_flg==3)
  429. save([save_path, 'trans_data_singlecon_KRR_10fold_normalized_allruns_set227_cliscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  430. elseif( nan_flg==4)
  431. save([save_path, 'trans_data_singlecon_KRR_10fold_normalized_allruns_set227_briefcomp_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  432. end
  434. end
  436. elseif( yeo_flg==1)
  437. if( scale_flg ==0)
  438. if( nan_flg==1)
  439. save([save_path, 'trans_data_yeonet_KRR_10fold_normalized_allruns_set227_cogbehav_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  440. elseif( nan_flg==2)
  441. save([save_path, 'trans_data_yeonet_KRR_10fold_normalized_allruns_set227_subscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  442. elseif( nan_flg==3)
  443. save([save_path, 'trans_data_yeonet_KRR_10fold_normalized_allruns_set227_clicscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  444. elseif( nan_flg==4)
  445. save([save_path, 'trans_data_yeonet_KRR_10fold_normalized_allruns_set227_briefcomp_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  446. end
  448. end
  449. elseif( yeo_flg==2)
  450. if( scale_flg ==0)
  451. if( nan_flg==1)
  452. save([save_path, 'trans_data_wholebrain_KRR_10fold_normalized_allruns_set227_cogbehav_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  453. elseif( nan_flg ==2)
  454. save([save_path, 'trans_data_wholebrain_KRR_10fold_normalized_allruns_set227_subscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  455. elseif( nan_flg==3)
  456. save([save_path, 'trans_data_wholebrain_KRR_10fold_normalized_allruns_set227_clicscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  457. elseif( nan_flg==4)
  458. save([save_path, 'trans_data_wholebrain_KRR_10fold_normalized_allruns_set227_briefcomp_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  459. end
  460. end
  462. elseif( yeo_flg==3)
  463. if( scale_flg ==0)
  464. if( nan_flg==1)
  465. save([save_path, 'trans_data_dmnEv_KRR_10fold_normalized_allruns_set227_cogbehav_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  466. elseif( nan_flg ==2)
  467. save([save_path, 'trans_data_dmnEv_KRR_10fold_normalized_allruns_set227_subscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  468. elseif( nan_flg==3)
  469. save([save_path, 'trans_data_dmnEv_KRR_10fold_normalized_allruns_set227_clicscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  470. elseif( nan_flg==4)
  471. save([save_path, 'trans_data_dmnEv_KRR_10fold_normalized_allruns_set227_briefcomp_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  472. end
  474. end
  476. elseif( yeo_flg==4)
  477. if( scale_flg ==0)
  478. if( nan_flg==1)
  479. save([save_path, 'trans_data_all6con_KRR_10fold_normalized_allruns_set227_cogbehav_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  480. elseif( nan_flg ==2)
  481. save([save_path, 'trans_data_all6con_KRR_10fold_normalized_allruns_set227_subscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  482. elseif( nan_flg==3)
  483. save([save_path, 'trans_data_all6con_KRR_10fold_normalized_allruns_set227_clicscores_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  484. elseif( nan_flg==4)
  485. save([save_path, 'trans_data_all6con_KRR_10fold_normalized_allruns_set227_briefcomp_lambda', num2str(ll)], 'all_perf_r', 'cur_lambda');
  486. end
  488. end
  489. end
  492. end

Testing_Set227_KRR.m at commit f57e5b7, no license · at the source

Overview

Authors: Anja Samardzija1, Xilin Shen2, Wenjing Luo1, Abigail S Greene3,4,5, Saloni Mehta2, Fuyuze Tokoglu2, Jagriti Arora2, Scott W Woods6, Rachel B Katz6, Gerard Sanacora6, Vinod H Srihari6, Dustin Scheinost1,2,3, R Todd Constable1,2,3
ORCID iDs: Anja Samardzija, Xilin Shen, Abigail S Greene, Saloni Mehta, Scott W Woods, Gerard Sanacora, Dustin Scheinost, R Todd Constable
  1. Department of Biomedical Engineering, Yale University, New Haven, CT USA
  2. Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT USA
  3. Interdepartmental Neuroscience Program, Yale School of Medicine, New Haven, CT USA
  4. MD–PhD program, Yale School of Medicine, New Haven, CT USA
  5. Department of Psychiatry, Brigham & Women’s Hospital, Boston, MA USA
  6. Department of Psychiatry, Yale School of Medicine, New Haven, CT USA
Institutions: Yale University (United States); Brigham and Women's Hospital (United States)
Journal: Nature communications, volume 17, issue 1, article 7650
Dates: received 16 December 2024; accepted 20 May 2026; published online 16 June 2026
Type: Research article · Language: English
License: CC BY-NC-ND
Identifiers: DOI 10.1038/s41467-026-73941-0 · PMID 42303975 · PMCID PMC13434263 · OpenAlex W7164878291
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: human (organism)
Methods: Statistics, Machine learning, Connectivity, fMRI & imaging
Keywords: Cognitive neuroscience, Computational neuroscience
MeSH: Brain*, Connectome*, Nerve Net*, Cognition, Female, Humans, Magnetic Resonance Imaging, Male, Models, Neurological (* major topic)
Topic: Chronic Disease Management Strategies (Epidemiology, Medicine), according to OpenAlex
Funding: NIMH NIH HHS (R01 MH121095); U.S. Department of Health & Human Services | NIH | National Institute of Mental Health (NIMH) (MH121095)
Citations: cited by 2 papers (Europe PMC); 105 references in the paper

Abstract

The abstract is not reproduced here: the paper's license (CC BY-NC-ND) does not allow it. Read it in the paper, at the publisher or on Europe PMC.

Repositories

Its files are read in the Code ↔ Paper reader above, with 6 matches between paragraphs and lines of code.

YaleMRRC/testing_the_tests

License: none: the authors keep all their rights
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Commit: f57e5b7eb0731255dcd3bfff45c0f4df9d3e0b4b, 14 April 2026
Languages: MATLAB (5)
Size: 24 files, 5 scripts
Software Heritage: not archived
Found in: “Code availability”
Holds: README, CITATION.cff
Not found: license file, environment file, tests, continuous integration, documentation
Availability: 1 check, the latest on 27 September 2026: the link answers
  • 27 September 2026: the link answers
6 files
At the source: github.com/YaleMRRC/testing_the_tests

Zenodo 19574108

License: CC-BY-4.0
State: the link answers, verified on 27 September 2026
Evidence: files inventoried
Size: 1 file
Software Heritage: not checked
Found in: the references
Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation
Availability: 1 check, the latest on 27 September 2026: the link answers (HTTP 200)
  • 27 September 2026: the link answers (HTTP 200)
6 files
At the source:

Code availability statement

The paper has a code availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

Read it in the paper: doi.org/10.1038/s41467-026-73941-0.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 10 scripts, each with its path and the digest of its content;
  • 6 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data availability statement

The paper has a data availability statement. Its license (CC BY-NC-ND) does not allow reproducing it here; in short, from what the harvester recognized in it:

  • no repository, dataset or request procedure was recognized in it

Read it in the paper: doi.org/10.1038/s41467-026-73941-0.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 27 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 13 authors, 2 keywords, 9 MeSH terms, 2 funders, 83 references.

Cite

This paper

Samardzija, A., Shen, X., Luo, W., Greene, A. S., Mehta, S., Tokoglu, F., Arora, J., Woods, S. W., Katz, R. B., Sanacora, G., Srihari, V. H., Scheinost, D., & Constable, R. T. (2026). Using connectome-based predictive models to reveal the systems standardized tests and clinical symptoms are reflecting. Nature communications, 17(1), 7650. https://doi.org/10.1038/s41467-026-73941-0

BibTeX

@article{samardzija2026using,
author = {Samardzija, Anja and Shen, Xilin and Luo, Wenjing and Greene, Abigail S and Mehta, Saloni and Tokoglu, Fuyuze and Arora, Jagriti and Woods, Scott W and Katz, Rachel B and Sanacora, Gerard and Srihari, Vinod H and Scheinost, Dustin and Constable, R Todd},
title = {{Using connectome-based predictive models to reveal the systems standardized tests and clinical symptoms are reflecting}},
journal = {Nature communications},
year = {2026},
month = jun,
volume = {17},
number = {1},
pages = {7650},
publisher = {Nature Publishing Group},
issn = {2041-1723},
doi = {10.1038/s41467-026-73941-0},
url = {https://doi.org/10.1038/s41467-026-73941-0},
pmid = {42303975},
pmcid = {PMC13434263}
}

RIS

TY - JOUR
AU - Samardzija, Anja
AU - Shen, Xilin
AU - Luo, Wenjing
AU - Greene, Abigail S
AU - Mehta, Saloni
AU - Tokoglu, Fuyuze
AU - Arora, Jagriti
AU - Woods, Scott W
AU - Katz, Rachel B
AU - Sanacora, Gerard
AU - Srihari, Vinod H
AU - Scheinost, Dustin
AU - Constable, R Todd
TI - Using connectome-based predictive models to reveal the systems standardized tests and clinical symptoms are reflecting
T2 - Nature communications
J2 - Nat Commun
PY - 2026
DA - 2026/06/16
VL - 17
IS - 1
SP - 7650
SN - 2041-1723
PB - Nature Publishing Group
DO - 10.1038/s41467-026-73941-0
UR - https://doi.org/10.1038/s41467-026-73941-0
LA - en
ER -

CSL-JSON

{
"id": "10.1038/s41467-026-73941-0",
"type": "article-journal",
"title": "Using connectome-based predictive models to reveal the systems standardized tests and clinical symptoms are reflecting",
"container-title": "Nature communications",
"author": [
{
"family": "Samardzija",
"given": "Anja"
},
{
"family": "Shen",
"given": "Xilin"
},
{
"family": "Luo",
"given": "Wenjing"
},
{
"family": "Greene",
"given": "Abigail S"
},
{
"family": "Mehta",
"given": "Saloni"
},
{
"family": "Tokoglu",
"given": "Fuyuze"
},
{
"family": "Arora",
"given": "Jagriti"
},
{
"family": "Woods",
"given": "Scott W"
},
{
"family": "Katz",
"given": "Rachel B"
},
{
"family": "Sanacora",
"given": "Gerard"
},
{
"family": "Srihari",
"given": "Vinod H"
},
{
"family": "Scheinost",
"given": "Dustin"
},
{
"family": "Constable",
"given": "R Todd"
}
],
"container-title-short": "Nat Commun",
"volume": "17",
"issue": "1",
"page": "7650",
"DOI": "10.1038/s41467-026-73941-0",
"PMID": "42303975",
"PMCID": "PMC13434263",
"ISSN": "2041-1723",
"publisher": "Nature Publishing Group",
"URL": "https://doi.org/10.1038/s41467-026-73941-0",
"language": "en",
"issued": {
"date-parts": [
[
2026,
6,
16
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1038/s44220-026-00623-7 [code]
Optimizing functional connectivity scanning conditions for predicting autistic traits.
Journal: Nature. Mental health
In common: 18 references, 3 authors
[2] doi:10.1038/s41562-026-02447-y [code]
Feature selection leads to divergent neurobiological interpretations of brain-based machine learning biomarkers.
Journal: Nature human behaviour
In common: 10 references, author Dustin Scheinost
[3] doi:10.7554/elife.104053 [code]
Brain cognition gaps reveal associations with dopamine and factors related to brain health through artificial intelligence prediction of functional connectome.
Journal: eLife
In common: 10 references
[4] doi:10.21203/rs.3.rs-9326213/v1 [code]
Multi-task fMRI outperforms resting-state fMRI for revealing task-invariant organization of the human brain
Journal: Research Square (preprint)
In common: 8 references
[5] doi:10.64898/2026.03.09.710558 [code]
Multi-task fMRI outperforms resting-state fMRI for revealing task-invariant organization of the human brain
Journal: bioRxiv (preprint)
In common: 8 references
[6] doi:10.1162/imag.a.1287 [code]
Is the whole more than the sum of its parts? Considering global and local features of the connectome improves prediction of individuals and phenotypes.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: 7 references
[7] doi:10.1038/s41467-026-75585-6 [code]
Brain network dynamics reflect psychiatric illness status and transdiagnostic symptom profiles across health and disease.
Journal: Nature communications
In common: 7 references
[8] doi:10.1002/hbm.70533 [code]
Trait-Relevant Tasks Improve Personality Prediction From Structural-Functional Brain Network Coupling.
Journal: Human brain mapping
In common: 6 references
[9] doi:10.1523/eneuro.0370-25.2026 [code]
Pretraining for Large-Scale Functional Connectome Fingerprinting Supports Generalization and Transfer Learning in Functional Neuroimaging.
Journal: eNeuro
In common: 7 references
[10] doi:10.1162/netn.a.549 [code]
Distributed cortical network dynamics of binocular convergent eye movements in humans.
Journal: Network neuroscience (Cambridge, Mass.)
In common: 7 references

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.