Alternating EM algorithm for a bilinear model in isoform quantification from RNA-seq data

Abstract

Motivation Estimation of isoform-level gene expression from RNA-seq data depends on simplifying assumptions, such as uniform read distribution, that are easily violated in real data. Such violations typically lead to biased estimates. Most existing methods provide bias correction step(s), which is based on biological considerations—such as GC content—and applied in single samples separately. The main problem is that not all biases are known. Results We have developed a novel method called XAEM based on a more flexible and robust statistical model. Existing methods are essentially based on a linear model Xβ, where the design matrix X is known and is computed based on the simplifying assumptions. In contrast XAEM considers Xβ as a bilinear model with both X and β unknown. Joint estimation of X and β is made possible by a simultaneous analysis of multi-sample RNA-seq data. Compared to existing methods, XAEM automatically performs empirical correction of potentially unknown biases. We use an alternating expectation-maximization (AEM) algorithm, alternating between estimation of X and β. For speed XAEM utilizes quasi-mapping for read alignment, thus leading to a fast algorithm. Overall XAEM performs favorably compared to recent advanced methods. For simulated datasets, XAEM obtains higher accuracy for multiple-isoform genes. In a differential-expression analysis of a real single-cell RNA-seq dataset, XAEM achieves substantially better rediscovery rates in independent validation sets. Availability and implementation The method and pipeline are implemented as a tool and freely available for use at http://fafner.meb.ki.se/biostatwiki/xaem/. Supplementary information Supplementary data are available at Bioinformatics online.

The transcriptome‐wide landscape of molecular subtype‐specific mRNA expression profiles in acute myeloid leukemia

“…The FASTQ files were obtained from the three cohorts and RNA‐seq reads were aligned to the human genome hg19. Also, XAEM 12 was used to obtain the gene expression (transcript per million ‐ TPM) from the RNA‐seq data. The calculation process of XAEM followed the instructions provided at http://fafner.meb.ki.se/biostatwiki/xaem/.…”

Isoform-level quantification for single-cell RNA sequencing

“…The joint estimation procedure is called an AEM algorithm, for which the exact formulas are given in Deng et al (2020) . At convergence, the output β represents the estimated transcript abundances for individual cells.…”

“…The key methodological innovation of Scasa is the in silico construction of the TCs, each with a corresponding initial design matrix X that adapts to the actual sequencing protocol used. As discussed previously ( Deng et al , 2020 ), X also automatically accounts for unknown biases in a sequencing protocol. Moreover, an explicitly available X makes the statistical processing of the paralogs (isoforms with highly similar sequences) tractable.…”

“…We used the approach of Deng et al (2020) to merge highly similar isoforms into isoform paralogs. Theoretically, the number of non-zero singular values of X determines the number of estimable paralogs; in practice, we set a low non-zero threshold (1/30).…”

Algorithms meet sequencing technologies – 10th edition of the RECOMB-Seq workshop

“…The XAEM method ( Deng et al., 2019 ) adopts a bilinear model for transcript-level quantification that aims to perform multi-sample inference, considering evidence from multiple samples within the same RNA-seq experiment jointly when performing quantification. The model can be viewed as a generalization of more common transcript quantification models where the so-called “design” matrix is fixed, and inference solves for the maximum likelihood parameters under this design matrix and the observed sequences.…”