A PLATFORM FOR DETECTION OF FUSION GENES IN ALL FROM TARGET CAPTURE NEXT-GENERATION SEQUENCING OF DNA AND RNA.
(Abstract release date: 05/19/16)
EHA Library. Grioni A. 06/10/16; 133142; P154
Mr. Andrea Grioni
Contributions
Contributions
Abstract
Abstract: P154
Type: Poster Presentation
Presentation during EHA21: On Friday, June 10, 2016 from 17:15 - 18:45
Location: Poster area (Hall H)
Background
Acute Lymphoblastic Leukemia (ALL), the most common leukemia in childhood, is characterized by genomic alterations, such as chromosomal translocations typically affecting a limited number of recurrent genes (targets) with several variable partner genes (partners). Fusion genes can be targeted by chemotherapy and/or new drugs, therefore genomic profiling of ALL has the potential to identify important new prognostic markers and potentially druggable targets.
Aims
Herein, we researched and developed a bioinformatic solution, ‘BreakingPoint’, for the robust detection of fusion genes. This was implemented in the context of an expert evaluation of Target Capture Next-Generation Sequencing (NGS-TC) protocols both at the DNA and RNA level on ALL samples. The platform will eventually be evaluated for routine clinical diagnosis.
Methods
‘BreakingPoint’ has been developed to detect fusion genes from NGS-TC datasets from any sequencing technology and any biological material. Its distinctive feature is that it initially identifies all aligned reads indicating hypothetical breakpoints on the ‘targets’, then clusters these into longer (thus more informative) consensus sequences, which are then studied across the genome to identify the unknown ‘partners’. We tested ‘BreakingPoint’ on NGS-TC datasets of: (i) DNA material from 10 patients with 7 known gene fusions (Illumina Nextera DNA capture); (ii) RNA material of 6 patients with 8 known gene fusions (NuGen Ovation RNA Fusion Panel); all on Illumina’s MiSeq platform. We additionally compared our method to the popular ‘Delly’ tool (DNA) and the ‘TopHat Alignment App’ on the ‘Illumina BaseSpace Cloud’ (RNA).
Results
On the DNA dataset, ‘BreakingPoint’ identified 6/7 known fusion genes (e.g. BCR/ABL1, PAX5/AUTS2) plus one not previously known. The missing P2RY8/CRLF2 gene fusion was most probably due to the low complexity of chromosome X affecting the in vitro capture and in silico alignment steps. Our method also detected deletions affecting the IKZF1 ‘target’, a gene prone to mutation in ALL patients. ‘Delly’ identified 5/7 known gene fusions, missing PAX5/AUTS2 and P2RY8/CRLF2. On the RNA dataset, ‘BreakingPoint’ identified all 8 known fusion genes, including P2RY8/CRLF2, plus a novel gene fusion (PAX5/ZCCHC7). Illumina’s ‘TopHat Alignment App’ detected 5/8 known gene fusions and the novel PAX5/ZCCHC7 rearrangement; false negatives (BCR/ABL1; MLL/AF4;P2RY8/CRLF2) were most likely caused by low read numbers affecting ‘targets’ coverage. ‘BreakingPoint’ false positives showed a common pattern of low sequence complexity and few representative reads (less than 2), and in many cases breakpoints fell in long intron non-coding regions (LINC), making them easily recognizable when compared to true positives. Finally, we have implemented a graphical user interface to run ‘BreakingPoint’ (and other plug-in methods, including ‘Delly’, if the user so wishes) and browse the results.
Conclusion
These results indicate that our approach sensitively and reliably detects fusion genes, irrespective of biological material or protocol, even in cases of low reads coverage. This already provides the potential to identify novel and conventional targetable fusion genes. Eventually, after further validation and evaluation, and coupled with the user-friendly interface, we believe that the ‘BreakingPoint’ platform can be introduced in routine clinical diagnosis.
Session topic: Acute lymphoblastic leukemia - Biology 1
Keyword(s): Acute lymphoblastic leukemia, Diagnosis, DNA damage, Prognostic factor
Type: Poster Presentation
Presentation during EHA21: On Friday, June 10, 2016 from 17:15 - 18:45
Location: Poster area (Hall H)
Background
Acute Lymphoblastic Leukemia (ALL), the most common leukemia in childhood, is characterized by genomic alterations, such as chromosomal translocations typically affecting a limited number of recurrent genes (targets) with several variable partner genes (partners). Fusion genes can be targeted by chemotherapy and/or new drugs, therefore genomic profiling of ALL has the potential to identify important new prognostic markers and potentially druggable targets.
Aims
Herein, we researched and developed a bioinformatic solution, ‘BreakingPoint’, for the robust detection of fusion genes. This was implemented in the context of an expert evaluation of Target Capture Next-Generation Sequencing (NGS-TC) protocols both at the DNA and RNA level on ALL samples. The platform will eventually be evaluated for routine clinical diagnosis.
Methods
‘BreakingPoint’ has been developed to detect fusion genes from NGS-TC datasets from any sequencing technology and any biological material. Its distinctive feature is that it initially identifies all aligned reads indicating hypothetical breakpoints on the ‘targets’, then clusters these into longer (thus more informative) consensus sequences, which are then studied across the genome to identify the unknown ‘partners’. We tested ‘BreakingPoint’ on NGS-TC datasets of: (i) DNA material from 10 patients with 7 known gene fusions (Illumina Nextera DNA capture); (ii) RNA material of 6 patients with 8 known gene fusions (NuGen Ovation RNA Fusion Panel); all on Illumina’s MiSeq platform. We additionally compared our method to the popular ‘Delly’ tool (DNA) and the ‘TopHat Alignment App’ on the ‘Illumina BaseSpace Cloud’ (RNA).
Results
On the DNA dataset, ‘BreakingPoint’ identified 6/7 known fusion genes (e.g. BCR/ABL1, PAX5/AUTS2) plus one not previously known. The missing P2RY8/CRLF2 gene fusion was most probably due to the low complexity of chromosome X affecting the in vitro capture and in silico alignment steps. Our method also detected deletions affecting the IKZF1 ‘target’, a gene prone to mutation in ALL patients. ‘Delly’ identified 5/7 known gene fusions, missing PAX5/AUTS2 and P2RY8/CRLF2. On the RNA dataset, ‘BreakingPoint’ identified all 8 known fusion genes, including P2RY8/CRLF2, plus a novel gene fusion (PAX5/ZCCHC7). Illumina’s ‘TopHat Alignment App’ detected 5/8 known gene fusions and the novel PAX5/ZCCHC7 rearrangement; false negatives (BCR/ABL1; MLL/AF4;P2RY8/CRLF2) were most likely caused by low read numbers affecting ‘targets’ coverage. ‘BreakingPoint’ false positives showed a common pattern of low sequence complexity and few representative reads (less than 2), and in many cases breakpoints fell in long intron non-coding regions (LINC), making them easily recognizable when compared to true positives. Finally, we have implemented a graphical user interface to run ‘BreakingPoint’ (and other plug-in methods, including ‘Delly’, if the user so wishes) and browse the results.
Conclusion
These results indicate that our approach sensitively and reliably detects fusion genes, irrespective of biological material or protocol, even in cases of low reads coverage. This already provides the potential to identify novel and conventional targetable fusion genes. Eventually, after further validation and evaluation, and coupled with the user-friendly interface, we believe that the ‘BreakingPoint’ platform can be introduced in routine clinical diagnosis.
Session topic: Acute lymphoblastic leukemia - Biology 1
Keyword(s): Acute lymphoblastic leukemia, Diagnosis, DNA damage, Prognostic factor
Abstract: P154
Type: Poster Presentation
Presentation during EHA21: On Friday, June 10, 2016 from 17:15 - 18:45
Location: Poster area (Hall H)
Background
Acute Lymphoblastic Leukemia (ALL), the most common leukemia in childhood, is characterized by genomic alterations, such as chromosomal translocations typically affecting a limited number of recurrent genes (targets) with several variable partner genes (partners). Fusion genes can be targeted by chemotherapy and/or new drugs, therefore genomic profiling of ALL has the potential to identify important new prognostic markers and potentially druggable targets.
Aims
Herein, we researched and developed a bioinformatic solution, ‘BreakingPoint’, for the robust detection of fusion genes. This was implemented in the context of an expert evaluation of Target Capture Next-Generation Sequencing (NGS-TC) protocols both at the DNA and RNA level on ALL samples. The platform will eventually be evaluated for routine clinical diagnosis.
Methods
‘BreakingPoint’ has been developed to detect fusion genes from NGS-TC datasets from any sequencing technology and any biological material. Its distinctive feature is that it initially identifies all aligned reads indicating hypothetical breakpoints on the ‘targets’, then clusters these into longer (thus more informative) consensus sequences, which are then studied across the genome to identify the unknown ‘partners’. We tested ‘BreakingPoint’ on NGS-TC datasets of: (i) DNA material from 10 patients with 7 known gene fusions (Illumina Nextera DNA capture); (ii) RNA material of 6 patients with 8 known gene fusions (NuGen Ovation RNA Fusion Panel); all on Illumina’s MiSeq platform. We additionally compared our method to the popular ‘Delly’ tool (DNA) and the ‘TopHat Alignment App’ on the ‘Illumina BaseSpace Cloud’ (RNA).
Results
On the DNA dataset, ‘BreakingPoint’ identified 6/7 known fusion genes (e.g. BCR/ABL1, PAX5/AUTS2) plus one not previously known. The missing P2RY8/CRLF2 gene fusion was most probably due to the low complexity of chromosome X affecting the in vitro capture and in silico alignment steps. Our method also detected deletions affecting the IKZF1 ‘target’, a gene prone to mutation in ALL patients. ‘Delly’ identified 5/7 known gene fusions, missing PAX5/AUTS2 and P2RY8/CRLF2. On the RNA dataset, ‘BreakingPoint’ identified all 8 known fusion genes, including P2RY8/CRLF2, plus a novel gene fusion (PAX5/ZCCHC7). Illumina’s ‘TopHat Alignment App’ detected 5/8 known gene fusions and the novel PAX5/ZCCHC7 rearrangement; false negatives (BCR/ABL1; MLL/AF4;P2RY8/CRLF2) were most likely caused by low read numbers affecting ‘targets’ coverage. ‘BreakingPoint’ false positives showed a common pattern of low sequence complexity and few representative reads (less than 2), and in many cases breakpoints fell in long intron non-coding regions (LINC), making them easily recognizable when compared to true positives. Finally, we have implemented a graphical user interface to run ‘BreakingPoint’ (and other plug-in methods, including ‘Delly’, if the user so wishes) and browse the results.
Conclusion
These results indicate that our approach sensitively and reliably detects fusion genes, irrespective of biological material or protocol, even in cases of low reads coverage. This already provides the potential to identify novel and conventional targetable fusion genes. Eventually, after further validation and evaluation, and coupled with the user-friendly interface, we believe that the ‘BreakingPoint’ platform can be introduced in routine clinical diagnosis.
Session topic: Acute lymphoblastic leukemia - Biology 1
Keyword(s): Acute lymphoblastic leukemia, Diagnosis, DNA damage, Prognostic factor
Type: Poster Presentation
Presentation during EHA21: On Friday, June 10, 2016 from 17:15 - 18:45
Location: Poster area (Hall H)
Background
Acute Lymphoblastic Leukemia (ALL), the most common leukemia in childhood, is characterized by genomic alterations, such as chromosomal translocations typically affecting a limited number of recurrent genes (targets) with several variable partner genes (partners). Fusion genes can be targeted by chemotherapy and/or new drugs, therefore genomic profiling of ALL has the potential to identify important new prognostic markers and potentially druggable targets.
Aims
Herein, we researched and developed a bioinformatic solution, ‘BreakingPoint’, for the robust detection of fusion genes. This was implemented in the context of an expert evaluation of Target Capture Next-Generation Sequencing (NGS-TC) protocols both at the DNA and RNA level on ALL samples. The platform will eventually be evaluated for routine clinical diagnosis.
Methods
‘BreakingPoint’ has been developed to detect fusion genes from NGS-TC datasets from any sequencing technology and any biological material. Its distinctive feature is that it initially identifies all aligned reads indicating hypothetical breakpoints on the ‘targets’, then clusters these into longer (thus more informative) consensus sequences, which are then studied across the genome to identify the unknown ‘partners’. We tested ‘BreakingPoint’ on NGS-TC datasets of: (i) DNA material from 10 patients with 7 known gene fusions (Illumina Nextera DNA capture); (ii) RNA material of 6 patients with 8 known gene fusions (NuGen Ovation RNA Fusion Panel); all on Illumina’s MiSeq platform. We additionally compared our method to the popular ‘Delly’ tool (DNA) and the ‘TopHat Alignment App’ on the ‘Illumina BaseSpace Cloud’ (RNA).
Results
On the DNA dataset, ‘BreakingPoint’ identified 6/7 known fusion genes (e.g. BCR/ABL1, PAX5/AUTS2) plus one not previously known. The missing P2RY8/CRLF2 gene fusion was most probably due to the low complexity of chromosome X affecting the in vitro capture and in silico alignment steps. Our method also detected deletions affecting the IKZF1 ‘target’, a gene prone to mutation in ALL patients. ‘Delly’ identified 5/7 known gene fusions, missing PAX5/AUTS2 and P2RY8/CRLF2. On the RNA dataset, ‘BreakingPoint’ identified all 8 known fusion genes, including P2RY8/CRLF2, plus a novel gene fusion (PAX5/ZCCHC7). Illumina’s ‘TopHat Alignment App’ detected 5/8 known gene fusions and the novel PAX5/ZCCHC7 rearrangement; false negatives (BCR/ABL1; MLL/AF4;P2RY8/CRLF2) were most likely caused by low read numbers affecting ‘targets’ coverage. ‘BreakingPoint’ false positives showed a common pattern of low sequence complexity and few representative reads (less than 2), and in many cases breakpoints fell in long intron non-coding regions (LINC), making them easily recognizable when compared to true positives. Finally, we have implemented a graphical user interface to run ‘BreakingPoint’ (and other plug-in methods, including ‘Delly’, if the user so wishes) and browse the results.
Conclusion
These results indicate that our approach sensitively and reliably detects fusion genes, irrespective of biological material or protocol, even in cases of low reads coverage. This already provides the potential to identify novel and conventional targetable fusion genes. Eventually, after further validation and evaluation, and coupled with the user-friendly interface, we believe that the ‘BreakingPoint’ platform can be introduced in routine clinical diagnosis.
Session topic: Acute lymphoblastic leukemia - Biology 1
Keyword(s): Acute lymphoblastic leukemia, Diagnosis, DNA damage, Prognostic factor
{{ help_message }}
{{filter}}