Statistical modeling of sequencing errors in SAGE libraries

Abstract

An implementation using R is distributed as an R package. An online version is available at http://tagcalling.mbgproject.org

Quantitative miRNA expression analysis: Comparing microarrays with next-generation sequencing

“…Also, sequence neighborhood-based correction methods developed for SAGE libraries (Akmaev and Wang 2004;Beissbarth et al 2004) may be adapted to correct sequence bias in next-generation sequencing data. However, although these algorithms may succeed somewhat in correcting sequencing bias, they are also susceptible to over-correction of low abundant miRNA sequences that are similar to high abundant reads with reduced sensitivity as a consequence.…”

Quality Control of Expression Profiling Data

“…Sequencing errors can still arise, and estimates of such errors can be generated. Statistical algorithms have been created to correct for these errors [57], but no all-inclusive solution to these problems have yet been reported for SAGE expression data. An ideal method should be able to indicate the degree of normalisation of a normalised library or for cancer staging.…”

Error Correction and Clustering Algorithms for Next Generation Sequencing

“…The model we propose is similar to that of RECOUNT [Wijaya et al, 2009] used to correct next generation short read counts. Both models derive from a method originally meant to detect sequencing errors in SAGE libraries [Beissbarth et al, 2004]. Our model differs from the previous models in that it works with kmers rather than full reads, since there is insufficient replication of full length reads in genomic data (as compared to transcriptome data).…”

“…The update equations are adapted from Beissbarth et al [2004] using a different error model and are given as follows:…”