Survival analysis has been used to characterize the time-to-event data. In medical studies, a typical application is to analyze the survival time of specific cancers by using high-dimensional gene expressions. The main challenges include the involvement of non-informaive gene expressions and possibly nonlinear relationship between survival time and gene expressions. Moreover, due to possibly imprecise data collection or wrong record, measurement error might be ubiquitous in the survival time and its censoring status. Ignoring measurement error effects may incur biased estimator and wrong conclusion. To tackle those challenges and derive a reliable estimation with efficiently computational implementation, we develop the R package AFFECT, which is referred to Accelerated Functional Failure time model with Error-Contaminated survival Times. This package aims to correct for measurement error effects in survival times and implements a boosting algorithm under corrected data to determine informative gene expressions as well as derive the corresponding nonlinear functions.