Limits...
No significant viral transcription detected in whole breast cancer transcriptomes.

Fimereli D, Gacquer D, Fumagalli D, Salgado R, Rothé F, Larsimont D, Sotiriou C, Detours V - BMC Cancer (2015)

Bottom Line: The presence of viral elements was confirmed in sample-matched exome sequences, but could not be confirmed by PCR or IHC.Our results show that no viral sequences are expressed in significant amounts in the BC investigated.The presence of non-transcribed viral DNA cannot be excluded.

View Article: PubMed Central - PubMed

Affiliation: IRIBHM - Université Libre de Bruxelles, ULB, Campus Erasme CP602, 808 route de Lennik, 1070, Brussels, Belgium. dfimerel@ulb.ac.be.

ABSTRACT

Background: Studies evaluating the presence of viral sequences in breast cancer (BC), including various strains of human papillomavirus and human herpes virus, have yielded conflicting results. Most were based on RT-PCR and in situ hybridization.

Methods: In this report we searched for expressed viral sequences in 58 BC transcriptomes using five distinct in silico methods. In addition, we complemented our RNA sequencing results with exome sequencing, PCR and immunohistochemistry (IHC) analyses. A control sample was used to test our in silico methods.

Results: All of the computational methods correctly detected viral sequences in the control sample. We identified a small number of viral sequences belonging to human herpesvirus 4 and 6 and Merkel cell polyomavirus. The extremely low expression levels-two orders of magnitude lower than in a typical hepatitis B virus infection in hepatocellular carcinoma-did not suggest active infections. The presence of viral elements was confirmed in sample-matched exome sequences, but could not be confirmed by PCR or IHC.

Conclusions: Our results show that no viral sequences are expressed in significant amounts in the BC investigated. The presence of non-transcribed viral DNA cannot be excluded.

Show MeSH

Related in: MedlinePlus

Computational pipelines for the detection of non-human sequences. Five different pipelines were used in this study. In pipelines 1, 2 and 3 raw reads were first aligned to the human genome and the unmapped reads were then used for alignment to the viral database with BWA, for alignment to the viral database with BLAST, and for de novo assembly and alignment to the viral database, respectively. In pipelines 4 and 5, raw reads were given as input to TriageTools and VirusSeq, respectively.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4374178&req=5

Fig1: Computational pipelines for the detection of non-human sequences. Five different pipelines were used in this study. In pipelines 1, 2 and 3 raw reads were first aligned to the human genome and the unmapped reads were then used for alignment to the viral database with BWA, for alignment to the viral database with BLAST, and for de novo assembly and alignment to the viral database, respectively. In pipelines 4 and 5, raw reads were given as input to TriageTools and VirusSeq, respectively.

Mentions: The four in-house and one published virus detection pipelines we implemented are depicted in Figure 1. Pipelines 1–3 and 5 first select reads that cannot be mapped to the human genome and attempt to map them on the RefSeq viral genome database and Gib-V database, respectively. In pipelines 1–3, reads with a viral hit are then screened against a comprehensive general-purpose sequence database (NCBI’s Non Redundant, NR) in order to further rule out their human origin. Reads with better BLAST hits to human than viral sequences are discarded. Pipelines 1–3 use the same initial alignment step and the same final filtering step (see Methods). However, they rely on different strategies for the viral sequences alignment. Pipeline 1 uses BWA, a standard short read aligner in NGS studies; pipeline 2 uses BLAST, a decade-old, proven aligner; pipeline 3 attempts to circumvent the limits of short reads by inserting a de novo assembly step before the viral screening. Pipeline 5 (VirusSeq) is conceptually similar to pipeline 1, but was set up by an independent team who took different technical routes at all steps of the implementation.Figure 1


No significant viral transcription detected in whole breast cancer transcriptomes.

Fimereli D, Gacquer D, Fumagalli D, Salgado R, Rothé F, Larsimont D, Sotiriou C, Detours V - BMC Cancer (2015)

Computational pipelines for the detection of non-human sequences. Five different pipelines were used in this study. In pipelines 1, 2 and 3 raw reads were first aligned to the human genome and the unmapped reads were then used for alignment to the viral database with BWA, for alignment to the viral database with BLAST, and for de novo assembly and alignment to the viral database, respectively. In pipelines 4 and 5, raw reads were given as input to TriageTools and VirusSeq, respectively.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4374178&req=5

Fig1: Computational pipelines for the detection of non-human sequences. Five different pipelines were used in this study. In pipelines 1, 2 and 3 raw reads were first aligned to the human genome and the unmapped reads were then used for alignment to the viral database with BWA, for alignment to the viral database with BLAST, and for de novo assembly and alignment to the viral database, respectively. In pipelines 4 and 5, raw reads were given as input to TriageTools and VirusSeq, respectively.
Mentions: The four in-house and one published virus detection pipelines we implemented are depicted in Figure 1. Pipelines 1–3 and 5 first select reads that cannot be mapped to the human genome and attempt to map them on the RefSeq viral genome database and Gib-V database, respectively. In pipelines 1–3, reads with a viral hit are then screened against a comprehensive general-purpose sequence database (NCBI’s Non Redundant, NR) in order to further rule out their human origin. Reads with better BLAST hits to human than viral sequences are discarded. Pipelines 1–3 use the same initial alignment step and the same final filtering step (see Methods). However, they rely on different strategies for the viral sequences alignment. Pipeline 1 uses BWA, a standard short read aligner in NGS studies; pipeline 2 uses BLAST, a decade-old, proven aligner; pipeline 3 attempts to circumvent the limits of short reads by inserting a de novo assembly step before the viral screening. Pipeline 5 (VirusSeq) is conceptually similar to pipeline 1, but was set up by an independent team who took different technical routes at all steps of the implementation.Figure 1

Bottom Line: The presence of viral elements was confirmed in sample-matched exome sequences, but could not be confirmed by PCR or IHC.Our results show that no viral sequences are expressed in significant amounts in the BC investigated.The presence of non-transcribed viral DNA cannot be excluded.

View Article: PubMed Central - PubMed

Affiliation: IRIBHM - Université Libre de Bruxelles, ULB, Campus Erasme CP602, 808 route de Lennik, 1070, Brussels, Belgium. dfimerel@ulb.ac.be.

ABSTRACT

Background: Studies evaluating the presence of viral sequences in breast cancer (BC), including various strains of human papillomavirus and human herpes virus, have yielded conflicting results. Most were based on RT-PCR and in situ hybridization.

Methods: In this report we searched for expressed viral sequences in 58 BC transcriptomes using five distinct in silico methods. In addition, we complemented our RNA sequencing results with exome sequencing, PCR and immunohistochemistry (IHC) analyses. A control sample was used to test our in silico methods.

Results: All of the computational methods correctly detected viral sequences in the control sample. We identified a small number of viral sequences belonging to human herpesvirus 4 and 6 and Merkel cell polyomavirus. The extremely low expression levels-two orders of magnitude lower than in a typical hepatitis B virus infection in hepatocellular carcinoma-did not suggest active infections. The presence of viral elements was confirmed in sample-matched exome sequences, but could not be confirmed by PCR or IHC.

Conclusions: Our results show that no viral sequences are expressed in significant amounts in the BC investigated. The presence of non-transcribed viral DNA cannot be excluded.

Show MeSH
Related in: MedlinePlus