CFMDS: CUDA-based fast multidimensional scaling for genome-scale data.

BACKGROUND: Multidimensional scaling (MDS) is a widely used approach to dimensionality reduction. It has been applied to feature selection and visualization in various areas. Among diverse MDS methods, the classical MDS is a simple and theoretically sound solution for projecting data objects onto a low dimensional space while preserving the original distances among them as much as possible. However, it is not trivial to apply it to genome-scale data (e.g., microarray gene expression profiles) on regular desktop computers, because of its high computational complexity. RESULTS: We implemented a highly-efficient software application, called CFMDS (CUDA-based Fast MultiDimensional Scaling), which produces an approximate solution of the classical MDS based on CUDA (compute unified device architecture) and the divide-and-conquer principle. CUDA is a parallel computing architecture exploiting the power of the GPU (graphics processing unit). The principle of divide-and-conquer was adopted for circumventing the small memory problem of usual graphics cards. Our application software has been tested on various benchmark datasets including microarrays and compared with the classical MDS algorithms implemented using C# and MATLAB. In our experiments, CFMDS was more than a hundred times faster for large data than such general solutions. Regarding the quality of dimensionality reduction, our approximate solutions were as good as those from the general solutions, as the Pearson's correlation coefficients between them were larger than 0.9. CONCLUSIONS: CFMDS is an expeditious solution for the data dimensionality reduction problem. It is especially useful for efficient processing of genome-scale data consisting of several thousands of objects in several minutes.

Cyclosporine and sirolimus pharmacokinetics and drug-to-drug interactions in kidney transplant recipients.

This study was conducted to evaluate the pharmacokinetics (pk) and drug interactions between cyclosporine (CsA) and sirolimus (SRL) in kidney transplant recipients. The morning (a.m.) and evening (p.m.) pk of CsA (4-5 mg/kg/dose) and SRL (2 mg, n = 20; 5 mg, n = 33) were evaluated on day 7 (n = 53). CsA showed circadian variation when comparing a.m. and p.m. administration [AUC: 8066 vs. 6699, P < 0.001 (CI 970.9; 1763.6); C0: 272 vs. 245, P = 0.007 (CI 7.5; 46.1)]. SRL showed dose-proportional pk. Significant and drug-to-drug concentration-dependent pk interactions were observed within a narrow concentration range for both drugs. A fivefold increase in SRL AUC (from a mean of 130 to 538 ng h/mL) was associated with a 25% increase in mean a.m. CsA AUC [7021 to 8811 ng h/mL, P = 0.037, CI (-3461.2; -118.9)] and with a 42% increase in mean p.m. CsA AUC [5386-7639, P = 0.024, CI (-4164.4; -340.7)]. A twofold increase in a.m. CsA AUC (from 5860 to 10 974 ng h/mL) was associated with a 70% increase in mean SRL AUC [223 to 380 ng h/mL, P = 0.0026, CI (-291.7; -22.8)]. A twofold increase in p.m. CsA AUC (from 4573 to 9692 ng h/mL) was associated with a 63% increase in mean SRL AUC [246 to 400 ng h/mL, P = 0.032, CI (-290.7; -16.6)]. CSA shows circadian pk regardless of sirolimus dose or blood concentration. Significant drug-to-drug interactions occur within narrow blood drug concentrations. The magnitude of the effect of CsA on SRL blood concentration is higher than that of SRL on CsA blood concentrations. These findings emphasize the need for therapeutic drug monitoring using this drug combination.

Tacrolimus pharmacokinetic drug interactions: effect of prednisone, mycophenolic acid or sirolimus.

This study was conducted to evaluate time-dependent pharmacokinetic changes and drug interactions over the first 6 months after transplantation in kidney transplant recipients receiving tacrolimus (TAC), prednisone (PRED) and mycophenolate mofetil (MMF) or sirolimus (SRL). Pharmacokinetic assessments were carried out at day 7 and months 1, 3, and 6 in kidney transplant recipients receiving TAC plus PRED with either MMF (2 g/day, n = 13) or SRL (15 mg loading dose, 5 mg for 7 days followed by 2 mg/day, n = 12). There were no differences in the main demographic characteristics or in mean PRED doses during the first 6 months after transplant. From day 7 to month 6, there was a 65% increase in TAC dose corrected exposure (dose corrected area under the curve; AUC) in patients receiving MMF (P = 0.005) and a 59% increase in TAC dose corrected exposure in patients receiving SRL (P = 0.008). From day 7 to month 6, there was a 72% increase in mycophenolate dose corrected exposure (P = 0.001) and a 65% increase in SRL dose corrected exposure (P = 0.008). TAC dose corrected exposure was 23% lower in patients receiving SRL compared with MMF (P = 0.012) on average over the study period. PRED dose reduction was associated with increase in TAC (in patients receiving SRL, P = 0.040) and mycophenolic acid (MPA) (P = 0.070) drug exposures. Tercile distribution of TAC drug exposure showed a positive correlation with mean SRL exposures (P = 0.016). Conversely, tercile distribution of SRL drug exposure showed a positive correlation with mean TAC exposures (P = 0.004). Time-dependent increases in TAC, MPA and SRL drug exposures occur up to 6 months after transplantation. Drug-to-drug interactions indicate that intense therapeutic drug monitoring is required to avoid under- or over-immunosuppression.

Circadian and time-dependent variability in tacrolimus pharmacokinetics.

Tacrolimus (TAC) is considered a critical dose drug. The purpose of our study was to investigate circadian and time-dependent changes in TAC pharmacokinetics over the first year after kidney transplantation. Pharmacokinetic (PK) studies were performed in 26 recipients of first living donor kidney transplants at day 7 after morning (a.m.) and evening (p.m.) doses of TAC. Additional serial PK studies were carried out in nine patients at month 6 (M6) and month 12 (M12). Blood samples were collected before 1, 1.5, 2, 2.5, 3, 4, 6, 8 and 12 h after TAC administration. Demographics, TAC and adjunctive immunosuppressive doses, hematology, and biochemistry were recorded in each PK study. Mean age was 37 years, body mass index 23 kg/m(2), 58% males, and 85% Caucasian. Higher AUC (231.4 vs. 220 ng.h/mL, P = 0.06) and C(max) (34.1 +/- 12.6 vs. 24.4 +/- 9.8 ng/mL, P < 0.001), and lower T(max) (1.6 +/- 0.8 vs. 2.7 +/- 2.0 h, P = 0.05) values were observed comparing a.m. and p.m. administrations. Comparing D7, M6 and M12, there was a significant increase in dose-normalized AUC (31.4 +/- 22.2 vs. 50.1 +/- 33 vs. 39.2 +/- 24.4 ng.h/mL/mg, P = 0.005), C(max) (4.4 +/- 2.4 vs. 7.8 +/- 3.5 vs. 6.0 +/- 3.3 ng/mL/mg, P < 0.001) and T(max) (1.6 +/- 1.1 vs. 1.7 +/- 0.4 vs. 1.8 +/- 0.8 h, P = 0.006), respectively. Over the first year the intraindividual variability of dose-normalized AUC, C(max) and C(0) were 82%, 72%, and 90%, respectively. No significant changes were observed comparing inter-individual variability of dose-normalized AUC (21%, 24%, 33%), C(max) (46%, 45%, 55%), C(0) (49%, 83%, 81%) at D7, M6 and M12, respectively. We observed a good correlation between a.m. and p.m. TAC AUC (r(2) = 0.90) and C(0) (r(2) = 0.88). Tacrolimus pharmacokinetics display circadian variation suggesting a slower and delayed absorption phase at nighttime. Tacrolimus also showed time-dependent PK changes, suggesting an improvement in absorption during the first 6 months. Despite circadian variation we observed good correlations between a.m. and p.m. TAC AUC (r(2) = 0.90) and C(0) (r(2) = 0.88) and between C(0) and total daily TAC exposure (a.m. + p.m. AUC) suggesting that trough-guided therapeutic monitoring is still a reliable and simple strategy to optimize the clinical use of TAC.