Package 'skater'

Description

Some types of data or results are indexed by two identifiers in two different columns corresponding to data points for pairs of observations. E.g., you may have columns called id1 and id2 that index the tibble for all possible pairs of results between samples A, B, and C. If you attempt to join two tibbles with by=c("id1", "id2"), the join will fail if samples are flipped from one dataset to another. E.g., one tibble may have id1=A and id2=B while the other has id1=B and id2=A. This function ensures that id1 is alphanumerically first while id2 is alphanumerically second. See examples.

Usage

arrange_ids(.data, .id1, .id2)

Arguments

Value

A tibble with id1 and id2 rearranged alphanumerically.

Examples

d1 <- tibble::tribble(
  ~id1, ~id2, ~results1,
  "a",  "b",       10L,
  "a",  "c",       20L,
  "c",  "b",       30L
)
d2 <- tibble::tribble(
  ~id1, ~id2,  ~results2,
  "b",  "a",       101L,
  "c",  "a",       201L,
  "b",  "c",       301L
)
# Inner join fails because id1!=id2.
dplyr::inner_join(d1, d2, by=c("id1", "id2"))
# Arrange IDs
d1 %>% arrange_ids(id1, id2)
d2 %>% arrange_ids(id1, id2)
# Inner join
dplyr::inner_join(arrange_ids(d1, id1, id2), arrange_ids(d2, id1, id2), by=c("id1", "id2"))
# Recursively, if you had more than two tibbles
list(d1, d2) %>%
  purrr::map(arrange_ids, id1, id2) %>%
  purrr::reduce(dplyr::inner_join, by=c("id1", "id2"))

Description

Calculates accuracy and related metrics.

Usage

calc_accuracy(tabble)

Arguments

Details

Calculates accuracy, lower and upper bounds, the guessing rate and p-value of the accuracy vs. the guessing rate. This function is called by confusion_matrix, but if this is all you want, you can simply supply the table to this function.

Value

A tibble with the corresponding statistics

Author(s)

Michael Clark (see m-clark/confusion_matrix).

See Also

binom.test

Description

Given a frequency table of predictions versus target values, calculate numerous statistics of interest.

Usage

calc_stats(tabble, prevalence = NULL, positive, ...)

Arguments

Details

Used within confusion_matrix to calculate various confusion matrix metrics. This is called by confusion_matrix, but if this is all you want you can simply supply the table.

Suppose a 2x2 table with notation

The formulas used here are:

$Sensitivity = A/(A+C)$

$Specificity = D/(B+D)$

$Prevalence = (A+C)/(A+B+C+D)$

$Positive Predictive Value = (sensitivity * prevalence)/((sensitivity*prevalence) + ((1-specificity)*(1-prevalence)))$

$Negative Predictive Value = (specificity * (1-prevalence))/(((1-sensitivity)*prevalence) + ((specificity)*(1-prevalence)))$

$Detection Rate = A/(A+B+C+D)$

$Detection Prevalence = (A+B)/(A+B+C+D)$

$Balanced Accuracy = (sensitivity+specificity)/2$

$Precision = A/(A+B)$

$Recall = A/(A+C)$

$F1 = harmonic mean of precision and recall = (1+beta^2)*precision*recall/((beta^2 * precision)+recall)$

where beta = 1 for this function.

$False Discovery Rate = 1 - Positive Predictive Value$

$False Omission Rate = 1 - Negative Predictive Value$

$False Positive Rate = 1 - Specificity$

$False Negative Rate = 1 - Sensitivity$

$D' = qnorm(Sensitivity) - qnorm(1 - Specificity)$

$AUC ~= pnorm(D'/sqrt(2))$

See the references for discussions of the first five formulas. Abbreviations:

Positive Predictive Value: PPV

Negative Predictive Value: NPV

False Discovery Rate: FDR

False Omission Rate: FOR

False Positive Rate: FPR

False Negative Rate: FNR

Value

A tibble with (at present) columns for sensitivity, specificity, PPV, NPV, F1 score, detection rate, detection prevalence, balanced accuracy, FDR, FOR, FPR, FNR. For more than 2 classes, these statistics are provided for each class.

Note

Different names are used for the same statistics.

Sensitivity: True Positive Rate, Recall, Hit Rate, Power

Specificity: True Negative Rate

Positive Predictive Value: Precision

False Negative Rate: Miss Rate, Type II error rate, beta

False Positive Rate: Fallout, Type I error rate, alpha

This function is called by confusion_matrix, but if this is all you want, you can simply supply the table to this function.

Author(s)

Michael Clark (see m-clark/confusion_matrix).

References

Kuhn, M. (2008), "Building predictive models in R using the caret package, " Journal of Statistical Software, (https://www.jstatsoft.org/article/view/v028i05).

Altman, D.G., Bland, J.M. (1994) "Diagnostic tests 1: sensitivity and specificity", British Medical Journal, vol 308, 1552.

Altman, D.G., Bland, J.M. (1994) "Diagnostic tests 2: predictive values," British Medical Journal, vol 309, 102.

Velez, D.R., et. al. (2008) "A balanced accuracy function for epistasis modeling in imbalanced datasets using multifactor dimensionality reduction.," Genetic Epidemiology, vol 4, 306.

Description

Given a vector of predictions and target values, calculate numerous statistics of interest. Modified from m-clark/confusion_matrix.

Usage

confusion_matrix(
  prediction,
  target,
  positive = NULL,
  prevalence = NULL,
  dnn = c("Predicted", "Target"),
  longer = FALSE,
  ...
)

Arguments

Details

This returns accuracy, agreement, and other statistics. See the functions below to find out more. Originally inspired by the confusionMatrix function from the caret package.

Value

A list of tibble(s) with the associated statistics and possibly the frequency table as list column of the first element. If classes contain >1 numeric class and a single non-numeric class (e.g., "1", "2", "3", and "Unrelated", the RMSE of the reciprocal of the Targets + 0.5 will also be returned.)

References

Kuhn, M., & Johnson, K. (2013). Applied predictive modeling.

See Also

calc_accuracy calc_stats

Examples

prediction = c(0,1,1,0,0,1,0,1,1,1)
target     = c(0,1,1,1,0,1,0,1,0,1)
confusion_matrix(prediction, target, positive = '1')

set.seed(42)
prediction = sample(letters[1:4], 250, replace = TRUE, prob = 1:4)
target     = sample(letters[1:4], 250, replace = TRUE, prob = 1:4)
confusion_matrix(prediction, target)

prediction = c(rep(1, 50), rep(2, 40), rep(3, 60))
target     = c(rep(1, 50), rep(2, 50), rep(3, 50))
confusion_matrix(prediction, target)
confusion_matrix(prediction, target) %>% purrr::pluck("Table")
confusion_matrix(prediction, target, longer=TRUE)
confusion_matrix(prediction, target, longer=TRUE) %>%
  purrr::pluck("Other") %>%
  tidyr::spread(Class, Value)

# Prediction with an unrelated class
prediction = c(rep(1, 50), rep(2, 40), rep(3, 60), rep("Unrelated", 55))
target     = c(rep(1, 50), rep(2, 50), rep(3, 55), rep("Unrelated", 50))
confusion_matrix(prediction, target)
# Prediction with two unrelated classes
prediction = c(rep(1, 50), rep(2, 40), rep("Third", 60), rep("Unrelated", 55))
target     = c(rep(1, 50), rep(2, 50), rep("Third", 55), rep("Unrelated", 50))
confusion_matrix(prediction, target)

Description

Creates a tibble with degree, expected kinship coefficient, and inference boundaries.

Rows will be created up to the max_degree, with an additional row for any relationship more distant than max_degree. The degree value for the final row will be NA. This represents inference criteria for "unrelated" individuals. See examples.

Usage

dibble(max_degree = 3L)

Arguments

Value

A tibble containing the degree, expected kinship coefficient (k), lower (l) and upper (u) inference bounds.

Examples

dibble(3)
dibble(10)

Description

Converts a PLINK-formatted fam file to a pedigree object using kinship2::pedigree.

Usage

fam2ped(fam)

Arguments

Value

A tibble with new listcol ped containing pedigrees from kinship2::pedigree.

Examples

famfile <- system.file("extdata", "3gens.fam", package="skater", mustWork=TRUE)
fam <- read_fam(famfile)
fam2ped(fam)

Description

This function is used to retrieve a relatedness measure from IBD segments. The relatedness value returned is the kinship coefficient.

Usage

ibd2kin(.ibd_data, .map, type = NULL)

Arguments

Details

The input data should be pairwise IBD segments prepared via read_ibd. The function will internally loop over each chromosome, and use a specified genetic map to convert shared segments to genetic units. After doing so, the function converts the shared length to a kinship coefficient by summing $0.5*IBD2 + 0.25*IBD1$.

Note that the data read in by read_ibd when source="pedsim" returns a list with separate tibbles for IBD1 and IBD2 segments. The current implementation of this function requires running this function independently on IBD1 and IBD2 segments, then summarizing (adding) the corresponding proportions. See examples.

Value

Tibble with three columns:

1. id1 (sample identifier 1)

2. id2 (sample identifier 2)

3. kinship (kinship coefficent derived from shared segments)

References

http://faculty.washington.edu/sguy/ibd_relatedness.html

Examples

pedsim_fp <- system.file("extdata", "GBR.sim.seg.gz", package="skater", mustWork=TRUE)
pedsim_seg <- read_ibd(pedsim_fp, source = "pedsim")
gmapfile <- system.file("extdata", "sexspec-avg-min.plink.map", package="skater", mustWork=TRUE)
gmap <- read_map(gmapfile)
ibd1_dat <- ibd2kin(.ibd_data=pedsim_seg$IBD1, .map=gmap, type="IBD1")
ibd2_dat <- ibd2kin(.ibd_data=pedsim_seg$IBD2, .map=gmap, type="IBD2")
dplyr::bind_rows(ibd1_dat,ibd2_dat) %>%
  dplyr::group_by(id1,id2) %>%
  dplyr::summarise(kinship = sum(kinship), .groups = "drop")

Description

This is an unexported helper used in in ibd2kin. The function interpolates over segments to apply genetic length to the segment. It is inspired by Python code distributed by the Browning lab (documentation).

Usage

interpolate(ibd_bp, chromgpos)

Arguments

Value

Numeric vector with the genetic distance shared at the segment.

References

http://faculty.washington.edu/sguy/ibd_relatedness.html

Description

"Converts" a kinship coefficient to put on the same scale as shared cM using the formula $cm <- pmin(3560, 4*pmax(0, k)*3560)$.

Usage

kin2cm(k)

Arguments

Value

A vector of numeric estimated cM, ranging from 0-3560.

References

https://dnapainter.com/tools/sharedcmv4.

https://www.ancestry.com/dna/resource/whitePaper/AncestryDNA-Matching-White-Paper.pdf.

https://verogen.com/wp-content/uploads/2021/03/snp-typing-uas-kinship-estimation-gedmatch-pro-tech-note-vd2020058-a.pdf.

Examples

kin2cm(.25)
kin2cm(.125)
kin2cm(.0625)
dibble(9) %>% dplyr::mutate(cm=kin2cm(k))

Description

Infers relationship degree given a kinship coefficient.

Usage

kin2degree(k, max_degree = 3L)

Arguments

Value

A vector with inferred degree, up to the maximum degree in dibble (anything more distant is NA, i.e., unrelated).

Examples

kin2degree(0.5)
kin2degree(0.25)
kin2degree(0.125)
kin2degree(0.0625)
kin2degree(0.03125)
kin2degree(0.03125, max_degree=5)
kin2degree(-0.05)
k <- seq(.02, .5, .03)
kin2degree(k)
kin2degree(k, max_degree=5)
tibble::tibble(k=k) %>% dplyr::mutate(degree=kin2degree(k))

Description

Converts a pedigree class object from fam2ped to a pairwise list of relationships and their expected/theoretical kinship coefficient.

Usage

ped2kinpair(ped)

Arguments

Value

A tibble containing all pairwise kinship coefficients from the input pedigree.

Examples

famfile <- system.file("extdata", "3gens.fam", package="skater", mustWork=TRUE)
famfile %>%
  read_fam() %>%
  fam2ped() %>%
  dplyr::mutate(kinpairs=purrr::map(ped, ped2kinpair)) %>%
  dplyr::select(fid, kinpairs) %>%
  tidyr::unnest(cols=kinpairs)

Description

Plot pedigree

Usage

plot_pedigree(ped, file = NULL, width = 10, height = 8)

Arguments

Value

No return value, called for side effects.

Description

Reads in an ⁠akt kin⁠ results file. Input file must have seven columns, whitespace delimited:

1. id1 (member 1)

2. id2 (member 2)

3. IBD0 (ratio of IBD0/All SNPS)

4. IBD1 (ratio of IBD1/All SNPS)

5. Kinship Coefficient

6. NSNPS

Usage

read_akt(file)

Arguments

Value

A tibble containing the 7 columns from the akt file.

Examples

aktFile <- system.file("extdata", "3gens.akt", package="skater", mustWork=TRUE)
akt <- read_akt(aktFile)
akt

Description

Reads in a PLINK-formatted .fam file. Input file must have six columns:

1. Family ID

2. Individual ID

3. Father ID

4. Mother ID

5. Sex

6. Affected Status

Usage

read_fam(file)

Arguments

Value

A tibble containing the 6 columns from the fam file.

Examples

famfile <- system.file("extdata", "3gens.fam", package="skater", mustWork=TRUE)
fam <- read_fam(famfile)
fam

Description

Reads in the inferred IBD segments from hapibd (documentation) or IBD segment file generated by ped-sim (documentation).

If reading a hapibd segment file, the input data should have the following columns:

1. First sample identifier

2. First sample haplotype index (1 or 2)

3. Second sample identifier

4. Second sample haplotype index (1 or 2)

5. Chromosome

6. Base coordinate of first marker in segment

7. Base coordinate of last marker in segment

8. cM length of IBD segment

If read a pedsim segment file, the input data should have the following columns:

1. First sample identifier

2. Second sample identifer

3. Chromosome

4. Physical position start

5. Physical position end

6. IBD type

7. Genetic position start

8. Genetic position end

9. Genetic length (end - start)

Usage

read_ibd(file, source)

Arguments

Value

if source="hapibd", a tibble is returned. If source="pedsim", a list with two tibble elements, IBD1 and IBD2 is returned. Both the hapibd tibble, and the two pedsim tibbles contain six columns:

1. id1 (sample identifier 1)

2. id2 (sample identifier 2)

3. chr (chromosome)

4. start (segment bp start coordinate)

5. end (segment bp end coordinate)

6. length (shared segment length in genetic units, cM)

References

https://github.com/browning-lab/hap-ibd#output-files

https://github.com/williamslab/ped-sim#output-ibd-segments-file

Examples

hapibd_fp <- system.file("extdata", "GBR.sim.ibd.gz", package="skater", mustWork=TRUE)
hapibd_seg <- read_ibd(hapibd_fp, source = "hapibd")
pedsim_fp <- system.file("extdata", "GBR.sim.seg.gz", package="skater", mustWork=TRUE)
pedsim_seg <- read_ibd(pedsim_fp, source = "pedsim")

Description

Reads in an ibis results file. Input file must have six columns, whitespace delimited:

1. id1 (member 1)

2. id2 (member 2)

3. Kinship Coefficient

4. IBD2 (ratio of IBD2/All SNPS)

5. Segment count

6. Kinship Degree

Usage

read_ibis(file)

Arguments

Value

A tibble containing the 6 columns from the ibis file.

Examples

ibisFile <- system.file("extdata", "3gens.ibis.coef", package="skater", mustWork=TRUE)
ibis <- read_ibis(ibisFile)
ibis

Description

This function reads in the content from a genetic map file to translate physical distance to genetic units (i.e. cM). Regardless of the source, the input file must be sex-averaged and in a tab-separated "Plink" format (documentation) with the following four columns and no header (i.e. no column names):

1. Chromosome

2. Identifier (ignored in read_map())

3. Length (genetic length within the physical position boundary)

4. Position (physical position boundary)

The columns must be in the order above. Note that only the first, third, and fourth columns are used in the function.

Usage

read_map(file)

Arguments

Details

The genetic map could come from different sources. One source is the HapMap map distributed by the Browning Lab (documentation). If this map file is used, the non-sex chromosomes can be downloaded and concatenated to a single file as follows:

wget https://bochet.gcc.biostat.washington.edu/beagle/genetic_maps/plink.GRCh37.map.zip
unzip plink.GRCh37.map.zip
cat *chr[0-9]*GRCh37.map | sort -k1,1 -k4,4 --numeric-sort > plink.allchr.GRCh37.map

Another source is a sex-specific map ("bherer") originally published by Bherer et al and recommended by the developers of ped-sim for simulating IBD segments (documentation). To retrieve and prep this map file for simulation:

# Get the refined genetic map and extract
wget --no-check-certificate https://github.com/cbherer/Bherer_etal_SexualDimorphismRecombination/raw/master/Refined_genetic_map_b37.tar.gz
tar xvfpz Refined_genetic_map_b37.tar.gz

# Format for ped-sim as per https://github.com/williamslab/ped-sim#map-file-
printf "#chr\tpos\tmale_cM\tfemale_cM\n" > sexspec.pedsim.map
for chr in {1..22}; do
  paste Refined_genetic_map_b37/male_chr$chr.txt Refined_genetic_map_b37/female_chr$chr.txt \
    | awk -v OFS="\t" 'NR > 1 && $2 == $6 {print $1,$2,$4,$8}' \
    | sed 's/^chr//' >> sexspec.pedsim.map;
done

# Clean up
rm -rf Refined_genetic_map_b37*

After this, the sexspec.pedsim.map file is ready for use in simulation. However, it must be averaged and reformatted to "Plink format" to use here:

cat sexspec.pedsim.map | grep -v "^#" | awk -v OFS="\t" '{print $1,".",($3+$4)/2,$2}' > sexspec-avg.plink.map

#' The genetic maps created above are in the tens of megabytes size range. This is trivial to store for most systems but a reduced version would increase portability and ease testing. This "minimum viable genetic map" could be used for testing and as installed package data in an R package for example analysis. Read more about minimum viable genetic maps at:

• Blog post: https://hapi-dna.org/2020/11/minimal-viable-genetic-maps/

• Github repo with python code: https://github.com/williamslab/min_map

The code as written below reduces the averaged sex-specific genetic map from 833776 to 28726 positions (~30X reduction!).

# Get minmap script from github
wget https://raw.githubusercontent.com/williamslab/min_map/main/min_map.py

# Create empty minmap
echo -n > sexspec-avg-min.plink.map

# For each autosome...
for chr in {1..22}; do
  echo "Working on chromosome $chr..."
  # First pull out just one chromosome
  grep "^${chr}[[:space:]]" sexspec-avg.plink.map > tmp.${chr}
  # Run the python script on that chromosome.
  # The genetic map column is 3rd column (2nd in 0-start). Physical position is last column (3 in 0-based)
  python3 min_map.py -mapfile tmp.${chr} -chr ${chr} -genetcol 2 -physcol 3 -noheader -error 0.05
  # Strip out the header and reformat back to plink format, and append to minmap file
  cat min_viable_map${chr}.txt | grep -v "^#" | awk -v OFS="\t" '{print $1,".",$4,$2}' >> sexspec-avg-min.plink.map
  # Clean up
  rm -f min_viable_map${chr}.txt tmp.${chr}
done

This averaged version of the Bherer sex-specific map, reduced to a minimum viable genetic map with at most 5% error, in Plink format, is available as installed package data (see examples). This is useful for testing code, but the full genetic map should be used for most analysis operations.

Value

A tibble containing 3 columns:

1. chr (chromosome)

2. value (genetic length within the physical position boundary)

3. bp (physical position boundary)

References

https://www.cog-genomics.org/plink/1.9/formats#map

https://bochet.gcc.biostat.washington.edu/beagle/genetic_maps/

https://github.com/williamslab/ped-sim#map-file

https://www.nature.com/articles/ncomms14994

https://www.nature.com/articles/ncomms14994

https://github.com/cbherer/Bherer_etal_SexualDimorphismRecombination

Examples

gmapfile <- system.file("extdata", "sexspec-avg-min.plink.map", package="skater", mustWork=TRUE)
gmap <- read_map(gmapfile)

Description

Reads in the output from plink2 --make-king-table (documentation). Input file must have six columns, tab delimited:

1. id1 (member 1)

2. id2 (member 2)

3. nsnps

4. hethet: proportion of sites where both are heterozygous

5. k: Kinship Coefficient

Usage

read_plink2_king(file)

Arguments

Value

A tibble containing the 6 columns from the plink2 --make-king-table output.

References

https://www.cog-genomics.org/plink/2.0/distance#make_king

Examples

plink2kingFile <- system.file("extdata", "plink2-king-table.tsv", package="skater", mustWork=TRUE)
plink2king <- read_plink2_king(plink2kingFile)
plink2king
plink2king %>% dplyr::filter(k>0.01)