PMC

Scanning HN-TROSY with COSCOMs overcomes shortfalls of strip matching. (a) HN-TROSY of the 37 kDa E_A with strips of HNH_pre (left) and HNN_pre (top) at the (H,N) coordinates of G195 (cyan), as well as strips of HNH_suc (bottom) and HNN_suc (right) at the coordinates of A194 (green). (b) Strip matching for the predecessor of G195. A194 was initially missing; its C^α was erroneously picked at the position indicated by the arrow. Correlations to A190 and A234 (very weak) are seen in HNH_pre and HNN_pre (unlabeled).

Spectral derivatives suppress spurious correlations in covariance NMR spectra. The ∗ and X indicate true and erroneous correlations, respectively: (a,b) at an index H(i) = a (see eqs and ); (c) _–1^* at an index H(i+1) = c*; (d) _–1^X for an erroneous correlation at H(i+1) = c^X. (e,f) Element-wise products of with _–1^* (v⃗*) and with _–1^X (v⃗^X). (g–j) Derivatives (′) of vectors in a–d, respectively. (k,l) Element-wise products of ′ with _–1^′* (v⃗^(′)*) and ′ with _–1^′X (v⃗^(′)X). v⃗^(′)* and v⃗^(′)X denote the products of the derivatives and not the derivatives of the products. (m) H(i+1) trace in HH_s at index H(i) = a, without derivatives. (n) Corresponding H(i+1) trace with derivatives.

Identification of unique proton sequential correlations when using spectral derivatives and when multiplying COSCOMs. (a,b) HNH_sca, (c,d) HNH_sco, and (e,f) HNH_scaco obtained by multiplying a and c and b and d, respectively. (a,c,e) Correlations obtained without derivatives in the carbon dimensions. (b,d,f) Correlations obtained with derivatives. Covariance was performed with the MATLAB covariance NMR toolbox. The amplitudes of signals labeled * are Σv⃗* and Σv⃗^(′)* in a and b, respectively, while those labeled X are Σv⃗^X and Σv⃗^(′)*X, with the vectors v⃗ as defined in Figure . The ∗ denotes the true correlation. Data collected with the 37 kDa E_A domain.

Differentiating between true sequential correlations (*) and erroneous correlations due to partially overlapping signals (X): (a) (C^α 1D trace) from HNCA at H(i) = 7.558 ppm and N(i) = 120.023 ppm; (b) _–1^* from 2D-H(NCO)CA at H(i+1) = 7.608 ppm; (c) _–1^X from 2D-H(NCO)CA at H(i+1) = 8.602 ppm. (d) Element-wise product of with _–1^* (v⃗*). (e) Element-wise product of with _–1^X (v⃗^X). (f–h) Derivatives of vectors in a–c, respectively. (i,j) Element-wise products of ′ with _–1^′* (v⃗^(′)*) and ′ with _–1^′X (v⃗^(′)X), respectively. The normalized sum of the elements of v⃗*, v⃗^(′)*, v⃗^X, and v⃗^(′)X lead to correlations that are highlighted by the symbols * and X in Figure a,b. Data collected with the 37 kDa E_A domain.