4-S-(propionic acid)sulfidocyclophosphamide

We applied our original algorithm OncoFinder [12 ] for the functional annotation of the primary expression data and for the calculation of the PAS scores. The extracted raw microarray expression data were quantile normalized according to [44 (link)]. Our approach to the transcriptome-wide gene expression analysis applies processing of these results with the signalome knowledge base developed by SABiosciences (http://www.sabiosciences.com/pathwaycentral.php). The algorithm utilizes a scheme that takes into account the overall impact of each gene product in the signaling pathway but ignores its position in the pathway graph. The formula used to calculate the pathway activation strength (PAS) for a given sample and a given pathway p is as follows:
$PA{S}_p={\displaystyle \sum _{n}AR{R}_{np}}\cdot BTI{F}_n\cdot \lg (CN{R}_n)$
Here the case-to-normal ratio, CNR_n, is the ratio of expression levels for a gene n in the sample under investigation to the same average value for the control group of samples. The Boolean flag of BTIF (beyond tolerance interval flag) equals zero when the CNR value has passed simultaneously the two criteria that demark the significantly perturbed expression level from essentially normal. The first criterion is the expression level for the sample lies within the tolerance interval, where p > 0.05. The second criterion is the discrete value of ARR (activator/repressor role) equals to the following fixed values: −1, when the gene/protein n is a repressor of pathway excitation; 1, if the gene/protein n is an activator of pathway excitation; 0, when the gene/protein n can be both an activator and a repressor of the pathway; 0.5 and −0.5, respectively, if the gene/protein n is rather an activator or repressor of the signaling pathway p, respectively. The results for the 82 pathways were obtained for each sample (listed in the Supplementary dataset 2). The area-under-curve (AUC) values were calculated according to [28 ]. Statistical tests were done using the R software package.

Data on Striga emergence count, ear rot, stalk and root lodging were transformed as [log (counts+1)] to reduce the heterogeneity of variances. The ANOVA for the 150 hybrids generated using the NCD II pooled over sets for each research condition [15 (link)] and across the stress conditions was carried out using the version 9.4 of SAS [33 ]. The genotypic component of the source of variation was partitioned into the variation due to males (sets), females (sets), and female × male (sets) interaction. The F-tests for male (sets), female (sets) and male × female (sets) mean squares were performed using male (sets) × environment, female (sets) × environment and male × female (sets) × environment mean squares, respectively. The mean squares attributable to environment × female × male (sets) were tested using the pooled error mean squares.
The following general linear model was used for the NCD II mating design:
$${\mathbf{X}}_{\mathbf{i}\mathbf{j}\mathbf{k}\mathbf{l}}=\mathbf{\mu }+{\mathbf{m}}_{\mathbf{i}}+{\mathbf{f}}_{\mathbf{i}}+{\left(\mathbf{m}\mathbf{f}\right)}_{\mathbf{i}\mathbf{j}}+{\mathbf{p}}_{\mathbf{i}\mathbf{j}\mathbf{k}}+{\mathbf{I}}_{\mathbf{l}}+{\mathbf{\epsilon }}_{\mathbf{i}\mathbf{j}\mathbf{k}\mathbf{l}}$$
where X_ijkl = the observed value of the progeny of the i^th male crossed with j^th female in the k^th replication; μ = the overall population mean; m_i = effect of the i^th female; f_j = the effect of the j^th male mated to the i^th female; (mf)_ij = the interaction effect between the i^thfemale and the j^th male; p_ijk = the effect of the k^th progeny from the cross between i^th female and j^th male; r_l = the effect of the l^th replication; ε_ijkl = the experimental error. The general combining ability (GCA) effects for male and female within sets (GCA_m and GCA_f) and specific combining ability (SCA) for each trait were estimated according to Kearsey and Pooni [34 ] as shown below:
$$\mathrm{G}\mathrm{C}{\mathrm{A}}_{\mathrm{m}}={\mathrm{X}}_{\mathrm{m}}-\mathrm{\mu }$$
$$\mathrm{G}\mathrm{C}{\mathrm{A}}_{\mathrm{f}}={\mathrm{X}}_{\mathrm{f}}-\mathrm{\mu }$$
where, GCA_m and GCA_f = General combining ability effects of male and female parents respectively; X_m and X_f = Average performance of a line when used as a male and female in crosses, respectively and μ = Overall mean of crosses in the set.
Standard errors (SE) for testing significance of GCA_m and GCA_f estimates, for traits of genotype, were computed from the mean squares of GCA_m × environment and GCA_f × environment, respectively as follows:
$$\mathrm{S}\mathrm{E}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{G}\mathrm{C}{\mathrm{A}}_{\mathrm{m}}=\surd \mathrm{M}\mathrm{S}\mathrm{m}\times \mathrm{e}/\left(\mathrm{f}\times \mathrm{e}\times \mathrm{r}\right)$$
$$\mathrm{S}\mathrm{E}\mathrm{f}\mathrm{o}\mathrm{r}\mathrm{G}\mathrm{C}{\mathrm{A}}_{\mathrm{f}}=\surd \mathrm{M}\mathrm{S}\mathrm{f}\times \mathrm{e}/\left(\mathrm{m}\times \mathrm{e}\times \mathrm{r}\right)$$
where, MSm × e and MSf × e were the mean squares of the interaction between male and environment as well as female × environment, respectively; f, m, r, and e were the number of females, males, replicates, and environments, respectively.
A multiple trait base index (MI) that integrated grain yield with the number of emerged Striga plants, Striga damage rating, plant and ear aspects, delayed leaf senescence, anthesis-silking interval and number of ears per plant was used to select the best performing hybrids across optimal, Striga and low-N conditions [5 (link)]. The means, adjusted for block effects of each genotype for each measured variable was standardized to minimize the effects of the different scales. A positive multiple trait base index value therefore indicated tolerance/resistance of the genotype to both Striga and low-N, while negative values indicated susceptibility to the stresses. The multiple trait base index was computed as follows:
MI = (2 × YLD) + EPP–EASP–PASP—STGR–RAT1 –RAT2 –(0.5 × C01)–(0.5 × C02)
On the other hand, the base indices for Striga and Low-N were computed as STRBI = 2.0 YLD + 1.0 EPP–(RAT1 + RAT2)– 0.5 (C01 + C02) and LNBI = 2.0 YLD + EPP–STGR–ASI—PASP–EASP, respectively to select superior hybrids under the respective stress conditions.
Where: MI = Multiple trait base index
STRBI = Base index for StrigaLNBI = Base index for Low-N
YLD = grain yield across research conditions
EPP = number of ears per plant across research conditions
EASP = Ear aspect across research conditions
PASP = Plant aspect across low-N and optimal conditions
STGR = Stay green characteristic across low-N conditions
RAT1 and RAT2 = Striga damage rating at 8 and 10 WAP across Striga infested conditions
C01 and C02 = Number of emerged Striga plants at 8 and 10 WAP across Striga -infested conditions.

All echocardiographic measurements were performed by two expert cardiologists (AP and GR) according to the instructions provided by the American Society of Echocardiography [16 (link)]. Therefore, principal measurements were recorded and independently reviewed by two distinct physicians. The systolic and diastolic volumes and ejection fraction were determined using apical two and four chamber views by Simpson biplane formula. We evaluated three consecutive cardiac cycles to obtain average pulsed Doppler transmitral flow velocity during early diastole velocity (E wave) and late diastole velocity (A wave) ratio (E/A) and the deceleration time (DT) of E. Placing the cursor laterally and medially at the mitral annulus level, we estimated mitral annulus movement by apical four-chamber tissue Doppler imaging (TDI). We obtained the recordings of systolic peak velocity (S’), early diastolic myocardial velocity (e’) and atrial systole velocity (A’), for three consecutive cardiac cycles. We calculated the ratio of peak early diastolic filling velocity and septal tissue Doppler early diastolic velocity (E/e’). In patients with atrial fibrillation, we measured E/e1 ratio and DT values. E/e’ > 15 was considered an index of elevated left ventricular (LV) filling pressure and severe diastolic dysfunction [17 (link)]. The tricuspid annular plane systolic excursion (TAPSE) was obtained by placing the M-mode cursor laterally to the tricuspid annulus. We estimated the PASP by continuous Doppler at tricuspid valve level, and PASP was obtained as the sum of the mean of 4 peak velocity of tricuspid regurgitation and the estimate of right atrial pressure based on ICV diameter and collapsibility. ICV dimension and collapse was calculated by subcostal view measuring in m-mode the internal vessel diameter soon before right atrium (RA) entrance during both expiration and inspiration. When the diameter reduction in inspiration was less than 50%, we added 10 mmHg to the TR velocity [18 (link)].