Limits...
Challenges in exome analysis by LifeScope and its alternative computational pipelines.

Pranckevičiene E, Rančelis T, Pranculis A, Kučinskas V - BMC Res Notes (2015)

Bottom Line: We summarized different approaches with regards to coverage (DP) and quality (QUAL) properties of the variants provided by GATK and found that LifeScope's computational pipeline is superior.We quantitatively supported a conclusion that Lifescope's pipeline is superior for processing sequencing data obtained by AB SOLiD 5500 system.It was noted that a coverage threshold for variant to be considered for further analysis has to be chosen in data-driven way to prevent a loss of important information.

View Article: PubMed Central - PubMed

Affiliation: Department of Human and Medical Genetics, Faculty of Medicine, Vilnius University, Santariskiu str. 2, LT-08661, Vilnius, Lithuania. erinija.pranckeviciene@mf.vu.lt.

ABSTRACT

Background: Every next generation sequencing (NGS) platform relies on proprietary and open source computational tools to analyze sequencing data. NGS tools for Illumina platforms are well documented which is not the case with AB SOLiD systems. We applied several computational and variant calling pipelines to analyse targeted exome sequencing data obtained using AB SOLiD 5500 system. Our investigated tools comprised proprietary LifeScope's pipeline in combination with open source color-space competent mapping programs and a variant caller. We present instrumental details of the pipelines that were used and quantitative comparative analysis of variant lists generated by LifeScope's pipeline versus open source tools.

Results: Sufficient coverage of targeted regions was achieved by all investigated pipelines. High variability was observed in identities of variants across the mapping programs. We observed less than 50% concordance of variant lists produced by approaches based on different mapping algorithms. We summarized different approaches with regards to coverage (DP) and quality (QUAL) properties of the variants provided by GATK and found that LifeScope's computational pipeline is superior. Fusion of information on mapping profiles (pileup) at genomic positions of variants in several different alignments proved to be a useful strategy to assess questionable singleton variants.

Conclusions: We quantitatively supported a conclusion that Lifescope's pipeline is superior for processing sequencing data obtained by AB SOLiD 5500 system. Nevertheless the use of alternative pipelines is encouraged because aggregation of information from other mapping and variant calling approaches helps to resolve questionable calls and increases the confidence of the call. It was noted that a coverage threshold for variant to be considered for further analysis has to be chosen in data-driven way to prevent a loss of important information.

No MeSH data available.


Related in: MedlinePlus

Decision tree diagram showing most discriminative annotations in classification of different categories of common variants identified in proband. Rectangular terminal nodes indicate fractions of variants (percentage) classified by the rules in each branch of the binary tree. Variable n indicates the number of samples from the training set assigned to that node. Node is associated with a class label of the most prevalent variant class. For example node 15 is associated with SHRiMP-GATK variants. Variant class identified by LifeScope-GATK is denoted by lg; SHRiMP-GATK is denoted by s; MAQ-GATK is denoted by m and BFAST-GATK by b. Tree nodes represented by ellipses show GATK variant annotations which were the most important in classifying the variants at each subsequent level. Classification rules are indicated by less or equal than and greater than conditions applied on a threshold value of the parameter. The diagram shows a considerable fraction of SHRiMP-GATK variants in node 15 characterized by larger depth of coverage (DP > 37). Large fraction of MAQ-GATK (m) variants are characterized by higher values of quality by depth (QD) (node 14) and have better GATK-assigned quality (see node 10). A group of LifeScope-GATK (lg) variants (node 14) are characterized by higher quality by depth (QD > 30.34). Another group (node 6) has lower QD and a negative value of BaseQRankSum, indicating poorer base quality support for alternative alleles. A tree size while running C50 algorithm was controlled constraining a split by minimum number of cases (parameter minCases) equal to 5000
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4562342&req=5

Fig5: Decision tree diagram showing most discriminative annotations in classification of different categories of common variants identified in proband. Rectangular terminal nodes indicate fractions of variants (percentage) classified by the rules in each branch of the binary tree. Variable n indicates the number of samples from the training set assigned to that node. Node is associated with a class label of the most prevalent variant class. For example node 15 is associated with SHRiMP-GATK variants. Variant class identified by LifeScope-GATK is denoted by lg; SHRiMP-GATK is denoted by s; MAQ-GATK is denoted by m and BFAST-GATK by b. Tree nodes represented by ellipses show GATK variant annotations which were the most important in classifying the variants at each subsequent level. Classification rules are indicated by less or equal than and greater than conditions applied on a threshold value of the parameter. The diagram shows a considerable fraction of SHRiMP-GATK variants in node 15 characterized by larger depth of coverage (DP > 37). Large fraction of MAQ-GATK (m) variants are characterized by higher values of quality by depth (QD) (node 14) and have better GATK-assigned quality (see node 10). A group of LifeScope-GATK (lg) variants (node 14) are characterized by higher quality by depth (QD > 30.34). Another group (node 6) has lower QD and a negative value of BaseQRankSum, indicating poorer base quality support for alternative alleles. A tree size while running C50 algorithm was controlled constraining a split by minimum number of cases (parameter minCases) equal to 5000

Mentions: Which of variant annotations discriminate the variant classes the best, was explored using common variants by C5.0 decision tree algorithm [28] in R. We do not attempt to fit a classification model but rather to perform exploratory analysis to discover thresholds of variant annotations best discriminating variant classes. C5.0 algorithm learns this information from data. A most discriminative annotation was root mean square mapping quality value (MQ). Almost all SHRiMP-GATK variants were assigned to a class characterized by high MQ (). Mapping quality computation is specific to each mapping program. Therefore, MQ values might not be directly comparable between the approaches. MQ and MQRankSum were exluded from the exploratory list of annotations and the additional sets of rules discriminating the variants were identified. Class of SHRiMP-GATK variants had a larger value of depth of coverage (DP). Class of MAQ-GATK variants was characterized by higher value of quality by depth (QD) and better quality assigned by GATK. One subclass of Lifescope-GATK variants had higher QD values. Another subclass of Lifescope-GATK variants had lower QD and a negative BaseQRankSum, indicating poorer base quality support for alternative alleles. The diagram of a decision tree of these classifications is shown in Fig. 5.


Challenges in exome analysis by LifeScope and its alternative computational pipelines.

Pranckevičiene E, Rančelis T, Pranculis A, Kučinskas V - BMC Res Notes (2015)

Decision tree diagram showing most discriminative annotations in classification of different categories of common variants identified in proband. Rectangular terminal nodes indicate fractions of variants (percentage) classified by the rules in each branch of the binary tree. Variable n indicates the number of samples from the training set assigned to that node. Node is associated with a class label of the most prevalent variant class. For example node 15 is associated with SHRiMP-GATK variants. Variant class identified by LifeScope-GATK is denoted by lg; SHRiMP-GATK is denoted by s; MAQ-GATK is denoted by m and BFAST-GATK by b. Tree nodes represented by ellipses show GATK variant annotations which were the most important in classifying the variants at each subsequent level. Classification rules are indicated by less or equal than and greater than conditions applied on a threshold value of the parameter. The diagram shows a considerable fraction of SHRiMP-GATK variants in node 15 characterized by larger depth of coverage (DP > 37). Large fraction of MAQ-GATK (m) variants are characterized by higher values of quality by depth (QD) (node 14) and have better GATK-assigned quality (see node 10). A group of LifeScope-GATK (lg) variants (node 14) are characterized by higher quality by depth (QD > 30.34). Another group (node 6) has lower QD and a negative value of BaseQRankSum, indicating poorer base quality support for alternative alleles. A tree size while running C50 algorithm was controlled constraining a split by minimum number of cases (parameter minCases) equal to 5000
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4562342&req=5

Fig5: Decision tree diagram showing most discriminative annotations in classification of different categories of common variants identified in proband. Rectangular terminal nodes indicate fractions of variants (percentage) classified by the rules in each branch of the binary tree. Variable n indicates the number of samples from the training set assigned to that node. Node is associated with a class label of the most prevalent variant class. For example node 15 is associated with SHRiMP-GATK variants. Variant class identified by LifeScope-GATK is denoted by lg; SHRiMP-GATK is denoted by s; MAQ-GATK is denoted by m and BFAST-GATK by b. Tree nodes represented by ellipses show GATK variant annotations which were the most important in classifying the variants at each subsequent level. Classification rules are indicated by less or equal than and greater than conditions applied on a threshold value of the parameter. The diagram shows a considerable fraction of SHRiMP-GATK variants in node 15 characterized by larger depth of coverage (DP > 37). Large fraction of MAQ-GATK (m) variants are characterized by higher values of quality by depth (QD) (node 14) and have better GATK-assigned quality (see node 10). A group of LifeScope-GATK (lg) variants (node 14) are characterized by higher quality by depth (QD > 30.34). Another group (node 6) has lower QD and a negative value of BaseQRankSum, indicating poorer base quality support for alternative alleles. A tree size while running C50 algorithm was controlled constraining a split by minimum number of cases (parameter minCases) equal to 5000
Mentions: Which of variant annotations discriminate the variant classes the best, was explored using common variants by C5.0 decision tree algorithm [28] in R. We do not attempt to fit a classification model but rather to perform exploratory analysis to discover thresholds of variant annotations best discriminating variant classes. C5.0 algorithm learns this information from data. A most discriminative annotation was root mean square mapping quality value (MQ). Almost all SHRiMP-GATK variants were assigned to a class characterized by high MQ (). Mapping quality computation is specific to each mapping program. Therefore, MQ values might not be directly comparable between the approaches. MQ and MQRankSum were exluded from the exploratory list of annotations and the additional sets of rules discriminating the variants were identified. Class of SHRiMP-GATK variants had a larger value of depth of coverage (DP). Class of MAQ-GATK variants was characterized by higher value of quality by depth (QD) and better quality assigned by GATK. One subclass of Lifescope-GATK variants had higher QD values. Another subclass of Lifescope-GATK variants had lower QD and a negative BaseQRankSum, indicating poorer base quality support for alternative alleles. The diagram of a decision tree of these classifications is shown in Fig. 5.

Bottom Line: We summarized different approaches with regards to coverage (DP) and quality (QUAL) properties of the variants provided by GATK and found that LifeScope's computational pipeline is superior.We quantitatively supported a conclusion that Lifescope's pipeline is superior for processing sequencing data obtained by AB SOLiD 5500 system.It was noted that a coverage threshold for variant to be considered for further analysis has to be chosen in data-driven way to prevent a loss of important information.

View Article: PubMed Central - PubMed

Affiliation: Department of Human and Medical Genetics, Faculty of Medicine, Vilnius University, Santariskiu str. 2, LT-08661, Vilnius, Lithuania. erinija.pranckeviciene@mf.vu.lt.

ABSTRACT

Background: Every next generation sequencing (NGS) platform relies on proprietary and open source computational tools to analyze sequencing data. NGS tools for Illumina platforms are well documented which is not the case with AB SOLiD systems. We applied several computational and variant calling pipelines to analyse targeted exome sequencing data obtained using AB SOLiD 5500 system. Our investigated tools comprised proprietary LifeScope's pipeline in combination with open source color-space competent mapping programs and a variant caller. We present instrumental details of the pipelines that were used and quantitative comparative analysis of variant lists generated by LifeScope's pipeline versus open source tools.

Results: Sufficient coverage of targeted regions was achieved by all investigated pipelines. High variability was observed in identities of variants across the mapping programs. We observed less than 50% concordance of variant lists produced by approaches based on different mapping algorithms. We summarized different approaches with regards to coverage (DP) and quality (QUAL) properties of the variants provided by GATK and found that LifeScope's computational pipeline is superior. Fusion of information on mapping profiles (pileup) at genomic positions of variants in several different alignments proved to be a useful strategy to assess questionable singleton variants.

Conclusions: We quantitatively supported a conclusion that Lifescope's pipeline is superior for processing sequencing data obtained by AB SOLiD 5500 system. Nevertheless the use of alternative pipelines is encouraged because aggregation of information from other mapping and variant calling approaches helps to resolve questionable calls and increases the confidence of the call. It was noted that a coverage threshold for variant to be considered for further analysis has to be chosen in data-driven way to prevent a loss of important information.

No MeSH data available.


Related in: MedlinePlus