Data used in this study were obtained from MCIC collection database at http:// schizconnect.org/ . This dataset was acquired utilizing a 1.5 T Siemens Sonata and DWI scans were collected. The imaging parameters were: TR = 9800 ms, TE = 86 ms, B values of 0 and 1000, NEX = 4, bandwidth = 1502, 64 slices, and 12 directions. Letter-number sequencing test (LNS) subtest of the Wechsler Adult Intelligence Scale—Third Edition 16 (WAIS-III) was applied as a complex verbal working memory task. The test involves a 24-item experimental condition, in which participants are read a series of letters and numbers and are asked to recite both back in ascending order, with the numbers first and then the letters. It is followed by a 24-item control condition that asks participants to simply repeat back the sequence of numbers and letters in the order presented.
Diffusion MRI Data Processing, Group Connectometry
Data analysis was conducted using DSI studio software available at dsi-studio. labsolver.org/ where instructions and technical illustrations are also provided. The diffusion data were reconstructed in the MNI space using q-space diffeomorphic reconstruction  to obtain the spin distribution function . A diffusion sampling length ratio of 1.25 was used, and the output resolution was 2 mm. Diffusion MRI connectometry  was conducted in a total of 29 patients using a multiple regression model considering letter-number sequencing test (LNS), sex, and age and the local connectomes expressing significant associations with the LNS were identified. The same approach was conducted in a total of 32 healthy controls matched for age and sex. Percentage thresholds of 30% to 50% were used to select local connectomes correlated with letter-number sequencing test for each group. A deterministic fiber tracking algorithm was conducted along the core pathway of fiber bundle to connect the selected local connectomes. A length threshold of 40 mm was used to select tracks. The seeding density was 20 seed(s) per mm3. To estimate the false discovery rate, a total of 2000 randomized permutations were applied to the group label to obtain the null distribution of the track length. Permutation testing allows for estimating and correcting the false discovery rate (FDR) of Type-I error inflation due to multiple comparisons.