triPOD

Sensitive detection of chromosomal abnormalities in parent-child trio-based microarray data.

Download .zip Download .tar.gz View on GitHub

triPOD

This software program detects chromosomal abnormalities in microarray data from parent-child trios using the Parent-of-Origin-based Detection (POD) method. It has been shown to provide improved sensitivity for abnormality detection, with substantial improvement in detection of low-level mosaicism, as described in Joseph D. Baugher, Benjamin D. Baugher, Matthew D. Shirley, and Jonathan Pevsner. Sensitive and specific detection of mosaic chromosomal abnormalities using the Parent-of-Origin-based Detection (POD) method. BMC Genomics 2013.

Requirements

triPOD is designed for use on Unix-based operating systems. It was developed and tested on Red Hat Enterprise Linux 6 using Perl v5.10.1 and R version 2.14.1.

Perl modules:

Algorithm::Cluster
threads
Tree::Interval

R libraries:

shape
TTR

Usage

perl triPOD.pl [--options] [INPUT_FILE]

Optional Arguments:
    --alpha   A threshold below which the familywise error rate is  
              controlled. Default = --alpha=0.1
    --batch   Submit a file containing a list of file names to be 
              analyzed.
    --build   The path to a file containing centromere locations from 
              the appropriate UCSC assembly.
              Default = --build=./genome_build/hg18_centromeres.txt
    --cite    Prints reference info for citations
    --cores   Number of CPU cores to employ 
              Default = maximum cores - 1 (e.g. --cores=8)
    --gender  Gender designation for sample (M or F). 
              Currently, chromosome X is analyzed only if a female 
              gender is specified. Default = NA
    --graph   Creates graphic output of results in PNG or PDF format 
              (--graph=none, --graph=png, --graph=pdf, or --graph=both)
              Default = --graph=none
    --hd      Abnormality detection by homozygous deletion analysis(PODhd) 
              (--hd or --nohd). Default = --hd
    --help    Prints a message describing options and input formatting
    --hetSD   Heterozygous SNP threshold as stdevs from the mean BAF
              Default = --hetSD=1.414213562373095 (sqrt of 2)
    --homSD   Homozygous SNP threshold as stdevs from the mean BAF 
              Default = --homSD=4
    --mi1     Abnormality detection by Mendelian "error" analysis(PODmi1) 
              (--mi1 or --nomi1). Default = --mi1
    --nc      Samples must have a No Call rate below this threshold to be 
              analyzed. Default = --nc=0.03
    --out     Specify an output directory (e.g. --out=results) 
              Default = --out=triPOD_Results
    --pod     Abnormality detection by standard POD algorithm 
              (--pod or --nopod). Default = --pod
    --podcr   Abnormality detection by cryptic POD algorithm (PODcr) 
              (--podcr or --nopodcr). Default = --podcr
    --stats   Create a file including calculated parameters, etc. 
              (--stats). Not created by default.
    --verbose Prints progress info to screen. Negated using --noverbose. 
              Default = --verbose.  
              Note that --batch mode will run in --noverbose mode.
    --win     Number of SNPs per window for sliding window analysis
              Default = --win=100

Note

  • It is highly recommended that parameters such as --alpha, --hetSD, --homSD, and --win are altered only by experienced users.

Input Formatting

The input file must be tab delimited, sorted by chromosome and position, and in the following order (columns): 

SNP_Name  Chromosome  Position  Father.GType  Father.BAF  Father.LRR  
  Mother.GType  Mother.BAF  Mother.LRR  Child.GType Child.BAF Child.LRR

A header line is expected and is used to extract sample names (i.e. Sample1.GType = Sample1), but not to determine column identity.

Genotype Annotation

The genotypes must be AA, AB, BB, NC (or NoCall).
triPOD has been developed using genotypes annotated by the Illumina method described in "TOP/BOT" Strand and "A/B" Allele. If converting from HapMap format (ATCG) to Illumina format (AB), use simple replacement as follows: 

AA = AA, TT; BB = CC, GG; AB = AC, AG, TC, TG; -- = NC.

Any markers with alternative genotyping combinations (e.g. CG or TA) should be discarded unless the user performs an additional analysis of the surrounding sequence.

B allele frequencies must be >= 0 and <= 1 for polymorphic markers.

Output

Several types of output are produced by triPOD.

  • *triPOD_Results.txt

    A tabular listing of detected abnormalities including the following information: Sample Name, Chromosome, Abnormal region start position, Abnormal region end position, Type of Abnormality, Parental Origin, Inheritance Pattern, Size of abnormal region (SNPs), Number of Informative SNPs, Size of abnormal region (bps), Detection Method, Descriptive statistics for the abnormal region: Child's normalized median mirrored BAF (mBAF) and LRR, Father's normalized median mirrored BAF (mBAF) and LRR, Mother's normalized median mirrored BAF (mBAF) and LRR.

    The results file also provides the date and time, triPOD version, the command line information, and the parameters employed.

  • *triPOD_Results.bed

    A .bed file is provided for visualization with genome browsers.

  • *triPOD_stats.txt

    An optional file which contains sample-specific calculations and parameters. See Baugher JD, et al. for explanations of rate and probability calculations. The following information is included for each trio: file name, trio member names, NoCall rates, nonadjacent homozygous deletion (HD) and single Mendelian error (MI1) rates, minimum region sizes, calculated MI1 BAF thresholds, initial and corrected alpha values for FDR control, estimated error rate, acceptable errors for POD region extension, global BAF thresholds for detecting informative SNPs, counts of regions detected by each detection method.

  • ./log_files/*triPOD_log.txt

    STDERR is redirected to a log file.

  • *.png and/or *.pdf

    Optional graphical output is generated for each chromosome harboring an abnormality. The upper panel is a plot of the log R ratio (LRR) values with the LRR moving average plotted in green and a black horizontal line at y=0 to assist with visualization of copy number changes. The center panel is a plot of the B allele frequency (BAF), in which abnormalities can frequently be visualized as splitting of the heterozygous (center) BAF band. The lower panel is a plot of the abnormalities detected by triPOD. Abnormalities with known parental origin or contribution are plotted adjacent to the Father and Mother axis labels. Abnormalities with unknown contribution or abnormal contribution from both parents are plotted along the y=0.5 axis, between the Father and Mother labels. The types of abnormalities are color coded as follows: red = deletion, blue = amplification, gray = homozygous deletion, purple = uniparental heterodisomy, green = uniparental isodisomy, and black = all other abnormalities for which a type cannot be assigned (including low-level mosaic events of all types).

Additional Considerations

  • Care should be taken when interpreting large abnormalities in commonly variable regions due to a tendency to combine small adjacent abnormalities of the same type.

  • Interesting findings should be graphically investigated until the user has gained expertise with the strengths and weaknesses of the various detection and annotation methods employed by triPOD.

Author

Joseph D. Baugher, Ph.D. - jbaugher(at)jhsph.edu