RT Journal Article
SR Electronic
T1 SMAP: exploiting high-throughput sequencing data of patient derived xenografts
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 440008
DO 10.1101/440008
A1 Yuna Blum
A1 Aurélien de Reyniès
A1 Nelson Dusetti
A1 Juan Iovanna
A1 Laetitia Marisa
A1 Rémy Nicolle
YR 2018
UL http://biorxiv.org/content/early/2018/10/11/440008.abstract
AB Background Patient-derived xenograft is the model of reference in oncology fordrug response analyses. Xenografts samples have the specificity to be composedof cells from both the graft and the host species. Sequencing analysis ofxenograft samples therefore requires specific processing methods to properlyreconstruct genomic profiles of both the host and graft compartments.Results We propose a novel xenograft sequencing process pipeline termedSMAP for Simultaneous mapping. SMAP integrates the distinction of host andgraft sequencing reads to the mapping process by simultaneously aligning to bothgenome references. We show that SMAP increases accuracy of species-assignmentwhile reducing the number of discarded ambiguous reads compared to otherexisting methods. Moreover, SMAP includes a module called SMAP-fuz toimprove the detection of chimeric transcript fusion in xenograft RNAseq data. Finally, we apply SMAP on a real dataset and show the relevance of pathway andcell population analysis of the tumoral and stromal compartments.Conclusions In high-throughput sequencing analysis of xenografts, our resultsshow that: i. the use of ad hoc sequence processing methods is essential, ii. highsequence homology does not introduce a significant bias when proper methodsare used and iii. the detection of fusion transcripts can be improved using ourapproach. SMAP is available on GitHub: cit-bioinfo.github.io/SMAP.