DNAfusion

https://github.com/CTrierMaansson/DNAfusion

Christoffer Trier Maansson1* and Emma Roger Andersen1

1Department of Clinical Biochemistry, Aarhus University Hospital, Denmark

*ctm@clin.au.dk

29 Oct 2024

Abstract

Circulating tumor DNA (ctDNA) containing somatic mutations can be found in blood plasma. This includes DNA fusions, such as the EML4-ALK, which can be an oncogenic driver in non-small cell lung cancer. This is an introduction to the DNAfusion package for R, which can be used to evaluate whether EML4-ALK is present in blood plasma.

1 Introduction
2 Installation
3 Package data
4 Functions
5 Session info

1 Introduction

This package was created in order to increase the sensitivity of EML4-ALK detection from commercially available NGS products such as the AVENIO (Roche) pipeline.

Paired-end sequencing of cfDNA generated BAM files can be used as input to discover EML4-ALK variants. This package was developed using position deduplicated BAM files generated with the AVENIO Oncology Analysis Software. These files are made using the AVENIO ctDNA surveillance kit and Illumina Nextseq 500 sequencing. This is a targeted hybridization NGS approach and includes ALK-specific but not EML4-specific probes.

The package includes eight functions.

The output of the first function, EML4_ALK_detection(), is used to determine whether EML4-ALK is detected and serves as input for the next four exploratory functions characterizing the EML4-ALK variant. The last function EML4_ALK_analysis() combines the output of the exploratory functions. The introns_ALK_EML4() function identifies the introns of EML4 and ALK containing the breakpoint. This is used in the find_variant() function which identifies the EML4-ALK variant.

To serve as examples, this package includes BAM files representing the EML4-ALK positive cell line H3122 and the EML4-ALK negative cell line, HCC827.

2 Installation

Use Bioconductor to install the most recent version of DNAfusion

if (!requireNamespace("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

BiocManager::install("DNAfusion")
library(DNAfusion)

3 Package data

BAM files from the cell lines, H3122 and HCC827, are included in the package and can be used as examples to explore the functions.

H3122_bam <- system.file("extdata", 
                            "H3122_EML4.bam",
                            package = "DNAfusion")
HCC827_bam <-  system.file("extdata", 
                            "HCC827_EML4.bam", 
                            package = "DNAfusion")

4 Functions

4.1 `EML4_ALK_detection()`

This function looks for EML4-ALK mate pair reads in the BAM file.

Input: \[\\[0.1in]\]

file

The name of the file which the data are to be read from.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can either be "hg38" or "hg19". 
Default = "hg38".

\[\\[0.1in]\]

mates

integer, the minimum number EML4-ALK mate pairs needed to be
detected in order to call a variant. Default = 2.

\[\\[0.1in]\]

Output:

A GAlignments object with soft-clipped reads representing EML4-ALK is returned. If no EML4-ALK is detected the the GAlignments is empty.

Examples:

H3122_result <- EML4_ALK_detection(file = H3122_bam, 
                        genome = "hg38", 
                        mates = 2) 
head(H3122_result)
#> GAlignments object with 6 alignments and 2 metadata columns:
#>       seqnames strand       cigar    qwidth     start       end     width
#>          <Rle>  <Rle> <character> <integer> <integer> <integer> <integer>
#>   [1]     chr2      +       94M2S        96  42299657  42299750        94
#>   [2]     chr2      +       92M4S        96  42299659  42299750        92
#>   [3]     chr2      +       92M4S        96  42299659  42299750        92
#>   [4]     chr2      +       91M5S        96  42299660  42299750        91
#>   [5]     chr2      +       91M5S        96  42299660  42299750        91
#>   [6]     chr2      +       87M9S        96  42299664  42299750        87
#>           njunc |      mpos                     seq
#>       <integer> | <integer>          <DNAStringSet>
#>   [1]         0 |  29223479 TTGCTTCTTT...GCAGTGGTCT
#>   [2]         0 |  29223606 GCTTCTTTCA...AGTGGTCTGA
#>   [3]         0 |  29223573 GCTTCTTTCA...AGTGGTCTGA
#>   [4]         0 |  29223650 CTTCTTTCAC...GTGGTCTGAT
#>   [5]         0 |  29223601 CTTCTTTCAC...GTGGTCTGAT
#>   [6]         0 |  29223640 TTTCACTTAG...TCTGATTTTT
#>   -------
#>   seqinfo: 455 sequences from an unspecified genome

HCC827_result <- EML4_ALK_detection(file = HCC827_bam, 
                    genome = "hg38", 
                    mates = 2)
HCC827_result
#> GAlignments object with 0 alignments and 0 metadata columns:
#>    seqnames strand       cigar    qwidth     start       end     width
#>       <Rle>  <Rle> <character> <integer> <integer> <integer> <integer>
#>        njunc
#>    <integer>
#>   -------
#>   seqinfo: no sequences

4.2 `EML4_sequence()`

This function identifies the basepairs leading up to the EML4 breakpoint.

Input: \[\\[0.1in]\]

reads

GAlignments object returned by EML4_ALK_detection().

\[\\[0.1in]\]

basepairs

integer, number of basepairs identified from the EML4-ALK fusion.
Default = 20.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can either be "hg38" or "hg19". 
Default = "hg38".

\[\\[0.1in]\]

Output:

If EML4-ALK is detected, returns a table of identified EML4 basepairs with the number of corresponding reads for each sequence. If no EML4-ALK is detected “No EML4-ALK was detected” is returned.

Examples:

EML4_sequence(H3122_result, genome = "hg38", basepairs = 20)
#> EML4_seq
#> CCAGGCTGGAGTGCAGTGGT GGAGTGCAGTGGTGTGATTT TCAGGCTGGAGTGCAGTGGT 
#>                   24                    1                    1
EML4_sequence(HCC827_result, genome = "hg38", basepairs = 20)
#> [1] "No EML4-ALK was detected"

4.3 `ALK_sequence()`

This function identifies the basepairs following the ALK breakpoint.

Input: \[\\[0.1in]\]

reads

GAlignments object returned by EML4_ALK_detection().

\[\\[0.1in]\]

basepairs

integer, number of basepairs identified from the EML4-ALK fusion. 
Default = 20.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can either be "hg38" or "hg19". 
Default = "hg38".

\[\\[0.1in]\]

Output:

If EML4-ALK is detected, returns a table of identified ALK basepairs with the number of corresponding reads for each sequence. If no spanning reads in ALK is detected an empty GAlignments object is returned. If no EML4-ALK is detected “No EML4-ALK was detected” is returned.

Examples:

ALK_sequence(H3122_result, genome = "hg38", basepairs = 20)
#> ALK_seq
#> AATGCAAAGCTAAAAATCAG ATGCAAAGCTAAAAATCAGA 
#>                    1                   36
ALK_sequence(HCC827_result, genome = "hg38", basepairs = 20)
#> [1] "No EML4-ALK was detected"

4.4 `break_position()`

This function identifies the genomic position in EML4 or ALK, where the breakpoint has happened.

Input: \[\\[0.1in]\]

reads

GAlignments object returned by EML4_ALK_detection().

\[\\[0.1in]\] genome

character representing the reference genome. 
Can either be "hg38" or "hg19". 
Default = "hg38".

\[\\[0.1in]\]

gene

Character string representing the gene. Can be either "ALK" or "EML4".

\[\\[0.1in]\]

Output:

If EML4-ALK is detected, it returns a table of genomic positions with the number of corresponding reads for each sequence. If no spanning reads in EML4 or ALK is detected an empty GAlignments object is returned. If no EML4-ALK is detected “No EML4-ALK was detected” is returned.

Examples:

break_position(H3122_result, genome = "hg38", gene = "EML4")
#> break_pos
#> 42299750 42299757 
#>       25        1
break_position(HCC827_result, genome = "hg38", gene = "EML4")
#> [1] "No EML4-ALK was detected"

4.5 `break_position_depth()`

This function identifies the read depth at the basepair before the breakpoint in EML4 or ALK.

Input: \[\\[0.1in]\]

file

The name of the file which the data are to be read from.

\[\\[0.1in]\]

reads

GAlignments returned by EML4_ALK_detection().

\[\\[0.1in]\] genome

character representing the reference genome. 
Can either be "hg38" or "hg19". 
Default = "hg38".

\[\\[0.1in]\]

gene

Character string representing the gene. Can be either "ALK" or "EML4".

\[\\[0.1in]\]

Output:

If EML4-ALK is detected a single integer corresponding to the read depth at the breakpoint is returned. If no spanning reads in EML4 or ALK is detected an empty GAlignments object is returned. If no EML4-ALK is detected “No EML4-ALK was detected” is returned.

Examples:

break_position_depth(H3122_bam, H3122_result, genome = "hg38", gene = "EML4")
#> [1] 30
break_position_depth(HCC827_bam, HCC827_result, genome = "hg38", gene = "EML4")
#> [1] "No EML4-ALK was detected"

4.6 `EML4_ALK_analysis()`

This functions collects the results from the other functions of the package.

Input: \[\\[0.1in]\]

file

The name of the file which the data are to be read from.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can be either "hg38" or "hg19".
Default = "hg38".

\[\\[0.1in]\]

mates

integer, the minimum number EML4-ALK mate pairs needed to be detected in
order to call a variant. Default = 2.

\[\\[0.1in]\]

basepairs

integer, number of basepairs identified from the EML4-ALK fusion. 
Default = 20.

\[\\[0.1in]\]

Output:

A list object with clipped_reads corresponding to EML4_ALK_detection(), last_EML4 corresponding to EML4_sequence(), first_ALK corresponding to ALK_sequence(), breakpoint_ALK corresponding to break_position(), gene = “ALK”, breakpoint_EML4 corresponding to break_position(), gene = “EML4”, read_depth_ALK corresponding to break_position_depth().gene = “ALK”, and read_depth_EML4 corresponding to break_position_depth() gene = “EML4”. If no EML4-ALK is detected an empty GAlignments is returned.

Examples:

H3122_results <- EML4_ALK_analysis(file = H3122_bam, 
                                    genome = "hg38", 
                                    mates = 2, 
                                    basepairs = 20)
HCC827_results <- EML4_ALK_analysis(file = HCC827_bam, 
                                    genome = "hg38", 
                                    mates = 2, 
                                    basepairs = 20)

head(H3122_results$clipped_reads)
#> GAlignments object with 6 alignments and 2 metadata columns:
#>       seqnames strand       cigar    qwidth     start       end     width
#>          <Rle>  <Rle> <character> <integer> <integer> <integer> <integer>
#>   [1]     chr2      +       94M2S        96  42299657  42299750        94
#>   [2]     chr2      +       92M4S        96  42299659  42299750        92
#>   [3]     chr2      +       92M4S        96  42299659  42299750        92
#>   [4]     chr2      +       91M5S        96  42299660  42299750        91
#>   [5]     chr2      +       91M5S        96  42299660  42299750        91
#>   [6]     chr2      +       87M9S        96  42299664  42299750        87
#>           njunc |      mpos                     seq
#>       <integer> | <integer>          <DNAStringSet>
#>   [1]         0 |  29223479 TTGCTTCTTT...GCAGTGGTCT
#>   [2]         0 |  29223606 GCTTCTTTCA...AGTGGTCTGA
#>   [3]         0 |  29223573 GCTTCTTTCA...AGTGGTCTGA
#>   [4]         0 |  29223650 CTTCTTTCAC...GTGGTCTGAT
#>   [5]         0 |  29223601 CTTCTTTCAC...GTGGTCTGAT
#>   [6]         0 |  29223640 TTTCACTTAG...TCTGATTTTT
#>   -------
#>   seqinfo: 455 sequences from an unspecified genome


H3122_results$last_EML4
#> EML4_seq
#> CCAGGCTGGAGTGCAGTGGT GGAGTGCAGTGGTGTGATTT TCAGGCTGGAGTGCAGTGGT 
#>                   24                    1                    1

H3122_results$first_ALK
#> ALK_seq
#> AATGCAAAGCTAAAAATCAG ATGCAAAGCTAAAAATCAGA 
#>                    1                   36

H3122_results$breakpoint_ALK
#> break_pos
#> 29223740 29223741 
#>        1       36

H3122_results$breakpoint_EML4
#> break_pos
#> 42299750 42299757 
#>       25        1

H3122_results$read_depth_ALK
#> [1] 827

H3122_results$read_depth_EML4
#> [1] 30

HCC827_results
#> GAlignments object with 0 alignments and 0 metadata columns:
#>    seqnames strand       cigar    qwidth     start       end     width
#>       <Rle>  <Rle> <character> <integer> <integer> <integer> <integer>
#>        njunc
#>    <integer>
#>   -------
#>   seqinfo: no sequences

4.7 `introns_ALK_EML4()`

This function identifies the introns of ALK and EML4 where the breakpoint has happened.

Input: \[\\[0.1in]\] file

The name of the file which the data are to be read from.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can be either "hg38" or "hg19".
Default = "hg38".

\[\\[0.1in]\]

Output:

Adataframeof the ALK- and EML4-intron of the breakpoint is returned corresponding to the transcript ENST00000389048.8 for ALK and ENST00000318522.10 for EML4. If the breakpoint is not located in introns of ALK or EML4, “Breakpoint not located in intron of ALK” or “Breakpoint not located in intron of EML4” is returned. If no EML4-ALK is detected “No EML4-ALK was detected” is returned.

Examples:

introns_ALK_EML4(file=H3122_bam,genome="hg38")
#>   intron_ALK intron_EML4
#> 1         19          13
introns_ALK_EML4(file=HCC827_bam,genome="hg38")
#> [1] "No ALK-EML4 was detected"

4.8 `find_variants()`

This function identifies the EML4-ALK variants as defined by Zhang et al. 2021

Input: \[\\[0.1in]\] file

The name of the file which the data are to be read from.

\[\\[0.1in]\]

genome

character representing the reference genome. 
Can be either "hg38" or "hg19".
Default = "hg38".

\[\\[0.1in]\]

Output:

A dataframeof the EML4-ALK variant is returned. If no variant is detected, “No ALK-EML4 was detected” is returned. If the variant is not defined a list with identified introns with breakpoints is returned. If the breakpoint could not be identified in either of the genes a list with identified introns with breakpoints is returned.

Examples:

find_variants(file=H3122_bam,genome="hg38")
#>               Variant Intron_EML4 Intron_ALK
#> 1 Variant 1 (E13,A20)          13         19
find_variants(file=HCC827_bam,genome="hg38")
#> [1] "No ALK-EML4 was detected"

5 Session info

#> ─ Session info ───────────────────────────────────────────────────────────────
#>  setting  value
#>  version  R version 4.4.1 (2024-06-14)
#>  os       Ubuntu 24.04.1 LTS
#>  system   x86_64, linux-gnu
#>  ui       X11
#>  language (EN)
#>  collate  C
#>  ctype    en_US.UTF-8
#>  tz       America/New_York
#>  date     2024-10-29
#>  pandoc   2.7.3 @ /usr/bin/ (via rmarkdown)
#> 
#> ─ Packages ───────────────────────────────────────────────────────────────────
#>  package   * version date (UTC) lib source
#>  BiocStyle * 2.34.0  2024-10-29 [2] Bioconductor 3.20 (R 4.4.1)
#>  DNAfusion * 1.8.0   2024-10-29 [1] Bioconductor 3.20 (R 4.4.1)
#> 
#>  [1] /home/biocbuild/bbs-3.20-bioc/tmpdir/RtmpUbgBc6/Rinst3d434c5c718c37
#>  [2] /home/biocbuild/bbs-3.20-bioc/R/site-library
#>  [3] /home/biocbuild/bbs-3.20-bioc/R/library
#> 
#> ──────────────────────────────────────────────────────────────────────────────

DNAfusion

https://github.com/CTrierMaansson/DNAfusion

29 Oct 2024

Abstract

Contents

1 Introduction

2 Installation

3 Package data

4 Functions

4.1 EML4_ALK_detection()

4.2 EML4_sequence()

4.3 ALK_sequence()

4.4 break_position()

4.5 break_position_depth()

4.6 EML4_ALK_analysis()

4.7 introns_ALK_EML4()

4.8 find_variants()

5 Session info

4.1 `EML4_ALK_detection()`

4.2 `EML4_sequence()`

4.3 `ALK_sequence()`

4.4 `break_position()`

4.5 `break_position_depth()`

4.6 `EML4_ALK_analysis()`

4.7 `introns_ALK_EML4()`

4.8 `find_variants()`