Impact assessment of the rational selection of training and test sets on the predictive ability of QSAR models.

This study performed an analysis of the influence of the training and test set rational selection on the quality and predictively of the quantitative structure-activity relationship (QSAR) model. The study was carried out on three different datasets of Influenza Neuraminidase (H1N1) inhibitors. The three datasets were divided into training and test sets using three rational selection methods: based on k-means, Kennard-Stone algorithm and Activity and the results were compared with Random selection. Then, a total of 31,490 mathematical models were developed and those models that presented a determination coefficient higher than: r2train > 0.8, r2loo > 0.7, r2test > 0.5 and minimum standard deviation (SD) and minimum root-mean square error (RMS) were selected. The selected models were validated using the internal leave-one-out method and the predictive capacity was evaluated by the external test set. The results indicate that random selection could lead to erroneous results. In return, a rational selection allows for obtaining more reliable conclusions. The QSAR models with major predictive power were found using the k-means algorithm and selection by activity.

[The monitoring of dialysis dose by ionic dialysance-based Kt reveals less dialysis adequacy than the Kt/V(UREA)-based measurement in critically ill patients with acute renal failure]. / La medida de la dosis de diálisis mediante Kt por dialisancia iónica revela una menor adecuación que la medida por Kt/V(UREA) en la insificiencia renal aguda de pacientes críticos.

INTRODUCTION: Measurement of dialysis dose by methods based on urea kinetics (Kt/VUREA) are hardly applicable to critical ill patients with acute renal failure (ARF). However, it is the base of the ADQI consensus recommendation for the target minimum dose. OBJECTIVE: To evaluate the usefulness of the real-time measurement of delivered dialysis dose (Kt) by means of the ionic dialysance (KtID) in the critically ill patient and to compare adequacy of dialysis dose between KtID and traditional Kt/V(UREA). MATERIAL AND METHODS: Prospective observational study in 17 critically ill patients with ARF requiring acute hemodialysis with a predefined prescription for the study (51 measures). RESULTS: The mean delivered Kt/V(UREA) was 1.19 +/- 0.14, with 59% of the sessions with values equal or above the ADQI recommendation. On the contrary, the mean KtID values obtained was 37.6 +/- 1 l, with only 29.4% of the sessions being equal or greater than the recommended values. CONCLUSIONS: Dialysis dose monitoring by means of KtID reveals a lower degree of adequacy as compared to the traditional Kt/V(UREA) method. The dynamic character of KtID monitoring can allow the adaptation of each dialysis session ("K" and/or "t") in order to achieve the recommended dose.