Dinh 2011/NOTES/2011-11-28

From ZhangLabWiki
Jump to navigation Jump to search

HL110 read 2 resynthesis[edit]

  • Using Read 2 Cluster Resynthesis Kit
  • Thaw content of kit on ice
  • Prepare HP3 (#19)
    • 2.85 mL PW1 + 150 ul of 2 N NaOH
  • Prepare HP2 (#16)
    • Content is 3 mL
    • Add: 15 ul of each: SolSeqV6.3.2r, N2RevSeq2, and Nextera2


Family background substraction with UCLA samples[edit]

  • The ANOVA is a generalization of the two sample t-test to find out whether multiple populations have equal means
    • If different families have different population means at a CpG site then it is influenced by family background
  • The ANOVA test requires a normality assumption, where the individual in each group is picked from a normal distribution
    • This is not true since there are affected/unaffected members in each family so it may be skewed.
  • The Kruskal-Wallis test can be applied to test where we test the null hypothesis that the samples come from populations such that the probability of a random observation from one group is greater than a random observation from another group is 0.5
    • Be careful using this test when the different groups have different shapes or variances.
  • The p value of the Kruskal-Wallis test follows a chi-square distribution, with degrees of freedom = the number of groups minus 1.
    • It does not follow chi-square if N in each group is < 5.
  • Script to perform test: File:CpgMatrixKrushalWallis.txt
    • Used only families with at least 5 members as groups to find the statistics.
    • Filter out sites where the p-value < 0.10

Hierarchical clustering[edit]

Example of how the h-clust were each generated

> A =read.table("sites_KW_signif0.10", header=TRUE, row.names=1)
> A$STDEV=NULL
> A$KW_pvalue=NULL
> A$min_RD=NULL
> A$mean_RD=NULL
> B=as.matrix(A)
> dissim=1-abs(cor(B,method="pearson", use="pairwise.complete.obs"))
> nrow(A)
> distance=as.dist(dissim)
> plot(hclust(distance, method="average"), main="Kruskal-Wallis significant sites, n=")
> pdf("Kruskal-Wallis significant sites 0.10")
> plot(hclust(distance, method="average"), main="Kruskal-Wallis significant sites, n=")
> dev.off()

Sample identity check[edit]

  • Took *.snp file generated by the mapping pipeline
  • Used bisSnpFilterHg19.pl to filter the SNPs
  • Count homozygous [ATGC] calls and heterozygous [^AGTC] calls for chromosome X
Comments Data_ID Number of heterozygous X SNP Number of homozygous X SNP %het %hom Sample_ID Sex
Mislabeled as female GK0210-002 2 62 3% 97% GK0210-002 Female
Mislabeled as female GK0277-002 2 45 4% 96% GK0277-002 Female
GK0178-003 57 70 45% 55% GK0178-003 Female
GK0100-003 47 48 49% 51% GK0100-003 Female
GK0050-003 54 55 50% 50% GK0050-003 Female
GK0050-002 54 54 50% 50% GK0050-002 Female
GK0100-002 67 65 51% 49% GK0100-002 Female
GK0277-004 58 56 51% 49% GK0277-004 Female
GK0344-003 68 65 51% 49% GK0344-003 Female
GK0271-003 70 66 51% 49% GK0271-003 Female
GK0359-004 59 54 52% 48% GK0359-004 Female
GK0005-002 76 69 52% 48% GK0005-002 Female
GK0123-005 62 56 53% 47% GK0123-005 Female
GK0167-006 56 50 53% 47% GK0167-006 Female
GK0122-003 51 45 53% 47% GK0122-003 Female
GK0091-003 72 63 53% 47% GK0091-003 Female
GK0391-002 54 46 54% 46% GK0391-002 Female
GK0271-004 74 62 54% 46% GK0271-004 Female
GK0365-002 62 51 55% 45% GK0365-002 Female
GK0123-003 64 51 56% 44% GK0123-003 Female
GK0241-004 51 40 56% 44% GK0241-004 Female
GK0167-003 48 37 56% 44% GK0167-003 Female
GK0295-002 55 42 57% 43% GK0295-002 Female
GK0396-002 75 57 57% 43% GK0396-002 Female
GK0187-002 62 47 57% 43% GK0187-002 Female
GK0281-002 59 43 58% 42% GK0281-002 Female
GK0027-003 66 46 59% 41% GK0027-003 Female
GK0359-003 63 42 60% 40% GK0359-003 Female
GK0241-002 76 50 60% 40% GK0241-002 Female
GK0290-002 64 42 60% 40% GK0290-002 Female
GK0005-004 0 53 0% 100% GK0005-004 Male
GK0091-001 0 64 0% 100% GK0091-001 Male
GK0123-001 0 56 0% 100% GK0123-001 Male
GK0178-004 0 77 0% 100% GK0178-004 Male
GK0187-001 0 83 0% 100% GK0187-001 Male
GK0277-005 0 62 0% 100% GK0277-005 Male
GK0281-001 0 51 0% 100% GK0281-001 Male
GK0290-001 0 65 0% 100% GK0290-001 Male
GK0290-004 0 60 0% 100% GK0290-004 Male
GK0295-003 0 64 0% 100% GK0295-003 Male
GK0344-002 0 70 0% 100% GK0344-002 Male
GK0344-004 0 53 0% 100% GK0344-004 Male
GK0391-001 0 70 0% 100% GK0391-001 Male
GK0005-001 1 80 1% 99% GK0005-001 Male
GK0271-001 1 79 1% 99% GK0271-001 Male
GK0365-003 1 79 1% 99% GK0365-003 Male
GK0391-003 1 77 1% 99% GK0391-003 Male
GK0210-003 1 75 1% 99% GK0210-003 Male
GK0277-003 1 72 1% 99% GK0277-003 Male
GK0123-004 1 71 1% 99% GK0123-004 Male
GK0344-001 1 70 1% 99% GK0344-001 Male
GK0396-003 1 70 1% 99% GK0396-003 Male
GK0187-004 1 69 1% 99% GK0187-004 Male
GK0100-001 1 67 1% 99% GK0100-001 Male
GK0281-003 1 65 2% 98% GK0281-003 Male
GK0100-006 1 64 2% 98% GK0100-006 Male
GK0122-001 1 63 2% 98% GK0122-001 Male
GK0295-004 1 63 2% 98% GK0295-004 Male
GK0396-001 1 62 2% 98% GK0396-001 Male
GK0100-004 1 61 2% 98% GK0100-004 Male
GK0178-002 1 60 2% 98% GK0178-002 Male
GK0359-006 1 60 2% 98% GK0359-006 Male
GK0167-002 1 59 2% 98% GK0167-002 Male
GK0241-005 1 59 2% 98% GK0241-005 Male
GK0122-004 1 56 2% 98% GK0122-004 Male
GK0365-001 1 56 2% 98% GK0365-001 Male
GK0359-001 1 54 2% 98% GK0359-001 Male
GK0005-003 1 51 2% 98% GK0005-003 Male
#1 GK0123-002 1 49 2% 98% GK0123-002 Male
GK0167-001 1 48 2% 98% GK0167-001 Male
GK0091-002 2 92 2% 98% GK0091-002 Male
GK0241-003 1 46 2% 98% GK0241-003 Male
GK0277-001 1 42 2% 98% GK0277-001 Male
GK0091-004 2 82 2% 98% GK0091-004 Male
GK0027-004 2 79 2% 98% GK0027-004 Male
GK0178-001 2 77 3% 97% GK0178-001 Male
GK0210-004 2 76 3% 97% GK0210-004 Male
GK0391-004 2 76 3% 97% GK0391-004 Male
GK0295-001 2 74 3% 97% GK0295-001 Male
GK0271-002 2 70 3% 97% GK0271-002 Male
GK0290-003 2 68 3% 97% GK0290-003 Male
GK0167-004 2 67 3% 97% GK0167-004 Male
GK0365-004 2 65 3% 97% GK0365-004 Male
GK0271-005 2 64 3% 97% GK0271-005 Male
GK0241-001 2 60 3% 97% GK0241-001 Male
GK0359-002 2 58 3% 97% GK0359-002 Male
GK0027-001 2 56 3% 97% GK0027-001 Male
GK0050-004 3 66 4% 96% GK0050-004 Male
#2 GK0123-002 2 44 4% 96% GK0123-002 Male
GK0027-002 3 63 5% 95% GK0027-002 Male
GK0122-002 3 60 5% 95% GK0122-002 Male
GK0187-003 3 57 5% 95% GK0187-003 Male
GK0050-001 4 63 6% 94% GK0050-001 Male
Mislabeled as male GK0396-004 62 70 47% 53% GK0396-004 Male
Mislabeled as male GK0210-001 63 59 52% 48% GK0210-001 Male
Mislabeled as male GK0050-005 63 57 53% 48% GK0050-005 Male
No data GK0281-004 XX XX XX XX GK0281-004 Male