TMRC2 Comprehensive Data Analysis: 202106

1 Introduction

This document is intended to provide a general overview of the TMRC2 samples which have thus far been sequenced. In some cases, this includes only those samples starting in 2019; in other instances I am including our previous (2015-2016) samples.

In all cases the processing performed was:

1. Default trimming was performed.
2. Hisat2 was used to map the remaining reads against the Leishmania panamensis genome revision 36.
3. The alignments from hisat2 were used to count reads/gene against the revision 36 annotations with htseq.
4. These alignments were also passed to the pileup functionality of samtools and the vcf/bcf utilities in order to make a matrix of all observed differences between each sample with respect to the reference.

The analyses in this document use the matrices of counts/gene from #3 and variants/position from #4 in order to provide some images and metrics describing the samples we have sequenced so far.

2 Annotations

Everything which follows depends on the Existing TriTrypDB annotations revision 46, circa 2019. The following block loads a database of these annotations and turns it into a matrix where the rows are genes and columns are all the annotation types provided by TriTrypDB.

The same database was used to create a matrix of orthologous genes between L.panamensis and all of the other species in the TriTrypDB.

tt <- sm(library(EuPathDB))
tt <- sm(library(org.Lpanamensis.MHOMCOL81L13.v46.eg.db))
pan_db <- org.Lpanamensis.MHOMCOL81L13.v46.eg.db
all_fields <- columns(pan_db)

all_lp_annot <- sm(load_orgdb_annotations(
    pan_db,
    keytype = "gid",
    fields = c("annot_gene_entrez_id", "annot_gene_name",
               "annot_strand", "annot_chromosome", "annot_cds_length",
               "annot_gene_product")))$genes

lp_go <- sm(load_orgdb_go(pan_db))
lp_lengths <- all_lp_annot[, c("gid", "annot_cds_length")]
colnames(lp_lengths)  <- c("ID", "length")
all_lp_annot[["annot_gene_product"]] <- tolower(all_lp_annot[["annot_gene_product"]])
orthos <- sm(EuPathDB::extract_eupath_orthologs(db = pan_db))

hisat_annot <- all_lp_annot
## rownames(hisat_annot) <- paste0("exon_", rownames(hisat_annot), ".E1")

3 TODO:

Resequence samples: TMRC20002, TMRC20006, TMRC20004 (maybe TMRC20008 and TMRC20029)

4 Generate Expressionsets and Sample Estimation

The process of sample estimation takes two primary inputs:

1. The sample sheet, which contains all the metadata we currently have on hand, including filenames for the outputs of #3 and #4 above.
2. The gene annotations.

An expressionset is a data structure used in R to examine RNASeq data. It is comprised of annotations, metadata, and expression data. In the case of our processing pipeline, the location of the expression data is provided by the filenames in the metadata.

The first lines of the following block create the Expressionset. All of the following lines perform various normalizations and generate plots from it.

4.1 Notes

The following samples are much lower coverage:

• TMRC20002
• TMRC20006
• TMRC20007
• TMRC20008

4.2 TODO:

1. Do the multi-gene family removal right here instead of way down at the bottom
2. Add zymodeme snps to the annotation later.
3. Start phylogenetic analysis of variant table.

sample_sheet <- glue::glue("sample_sheets/tmrc2_samples_20210601.xlsx")

sanitize_columns <- c("passagenumber", "clinicalresponse", "clinicalcategorical",
                      "zymodemecategorical", "phenotypiccharacteristics")
lp_expt <- sm(create_expt(sample_sheet,
                          gene_info = hisat_annot,
                          id_column = "hpglidentifier",
                          file_column = "lpanamensisv36hisatfile")) %>%
  set_expt_conditions(fact = "zymodemecategorical") %>%
  subset_expt(nonzero = 8600) %>%
  semantic_expt_filter(semantic = c("amastin", "gp63", "leishmanolysin"),
                       semantic_column = "annot_gene_product") %>%
  sanitize_expt_metadata(columns = sanitize_columns) %>%
  set_expt_factors(columns = sanitize_columns, class = "factor")

## The samples (and read coverage) removed when filtering 8600 non-zero genes are:

## TMRC20002 TMRC20004 TMRC20006 TMRC20029 TMRC20008 
##  11681227    564812   6670348   1658096   6249790

## subset_expt(): There were 52, now there are 47 samples.

## semantic_expt_filter(): Removed 68 genes.

libsizes <- plot_libsize(lp_expt)
libsizes$plot

## I think samples 7,10 should be removed at minimum, probably also 9,11
nonzero <- plot_nonzero(lp_expt)
nonzero$plot

## Warning: ggrepel: 6 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps

plot_boxplot(lp_expt)

## 3164 entries are 0.  We are on a log scale, adding 1 to the data.

filter_plot <- plot_libsize_prepost(lp_expt)
filter_plot$lowgene_plot

## Warning: Using alpha for a discrete variable is not advised.

filter_plot$count_plot

4.3 Distribution Visualization

Najib’s favorite plots are of course the PCA/TNSE. These are nice to look at in order to get a sense of the relationships between samples. They also provide a good opportunity to see what happens when one applies different normalizations, surrogate analyses, filters, etc. In addition, one may set different experimental factors as the primary ‘condition’ (usually the color of plots) and surrogate ‘batches’.

4.4 By Susceptilibity

Column ‘Q’ in the sample sheet, make a categorical version of it with these parameters:

• 0 <= x <= 35 is resistant
• 36 <= x <= 48 is ambiguous
• 49 <= x is sensitive

starting <- as.numeric(pData(lp_expt)[["susceptibilityinfectionreduction32ugmlsbvhistoricaldata"]])
sus_categorical <- starting
na_idx <- is.na(starting)
sus_categorical[na_idx] <- "unknown"

resist_idx <- starting <= 0.35
sus_categorical[resist_idx] <- "resistant"
indeterminant_idx <- starting >= 0.36 & starting <= 0.48
sus_categorical[indeterminant_idx] <- "ambiguous"
susceptible_idx <- starting >= 0.49
sus_categorical[susceptible_idx] <- "sensitive"

pData(lp_expt$expressionset)[["sus_category"]] <- sus_categorical

clinical_samples <- lp_expt %>%
  set_expt_batches(fact = sus_categorical)

clinical_norm <- sm(normalize_expt(clinical_samples, norm = "quant", transform = "log2",
                                   convert = "cpm", batch = FALSE, filter = TRUE))
zymo_pca <- plot_pca(clinical_norm, plot_title = "PCA of parasite expression values")
pp(file = "images/zymo_pca_sus_shape.png", image = zymo_pca$plot)

zymo_3dpca <- plot_3d_pca(zymo_pca)
zymo_3dpca$plot

zymo_tsne <- plot_tsne(clinical_norm, plot_title = "TSNE of parasite expression values")
zymo_tsne$plot

clinical_nb <- normalize_expt(clinical_samples, convert = "cpm", transform = "log2",
                         filter = TRUE, batch = "svaseq")

## Removing 145 low-count genes (8565 remaining).

## batch_counts: Before batch/surrogate estimation, 568 entries are x==0: 0%.

## batch_counts: Before batch/surrogate estimation, 2107 entries are 0<x<1: 1%.

## Setting 210 low elements to zero.

## transform_counts: Found 210 values equal to 0, adding 1 to the matrix.

clinical_nb_pca <- plot_pca(clinical_nb, plot_title = "PCA of parasite expression values")
pp(file = "images/clinical_nb_pca_sus_shape.png", image = clinical_nb_pca$plot)

clinical_nb_tsne <- plot_tsne(clinical_nb, plot_title = "TSNE of parasite expression values")
clinical_nb_tsne$plot

corheat <- plot_corheat(clinical_norm, plot_title = "Correlation heatmap of parasite
                 expression values
")
corheat$plot

plot_sm(clinical_norm)$plot

## Performing correlation.

4.5 By Cure/Fail status

cf_expt <- set_expt_conditions(lp_expt, fact = "clinicalcategorical") %>%
  set_expt_batches(fact = sus_categorical)

cf_norm <- normalize_expt(cf_expt, convert = "cpm", transform = "log2",
                          norm = "quant", filter = TRUE)

## Removing 145 low-count genes (8565 remaining).

## transform_counts: Found 2 values equal to 0, adding 1 to the matrix.

start_cf <- plot_pca(cf_norm, plot_title = "PCA of parasite expression values")
pp(file = "images/cf_sus_shape.png", image = start_cf$plot)

cf_nb <- normalize_expt(cf_expt, convert = "cpm", transform = "log2",
                        norm = "quant", filter = TRUE, batch = "svaseq")

## Warning in normalize_expt(cf_expt, convert = "cpm", transform = "log2", :
## Quantile normalization and sva do not always play well together.

## Removing 145 low-count genes (8565 remaining).

## batch_counts: Before batch/surrogate estimation, 2 entries are x==0: 0%.

## batch_counts: Before batch/surrogate estimation, 2573 entries are 0<x<1: 1%.

## Setting 103 low elements to zero.

## transform_counts: Found 103 values equal to 0, adding 1 to the matrix.

cf_nb_pca <- plot_pca(cf_nb, plot_title = "PCA of parasite expression values")
pp(file = "images/cf_sus_share_nb.png", image = cf_nb_pca$plot)

## Warning: ggrepel: 1 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps

cf_norm <- normalize_expt(cf_expt, transform = "log2", convert = "cpm",
                          filter = TRUE, norm = "quant")

## Removing 145 low-count genes (8565 remaining).

## transform_counts: Found 2 values equal to 0, adding 1 to the matrix.

test <- pca_information(cf_norm,
                        expt_factors = c("clinicalcategorical", "zymodemecategorical",
                                         "pathogenstrain", "passagenumber"),
                        num_components = 6, plot_pcas = TRUE)
test$anova_p

##                          PC1     PC2       PC3     PC4    PC5    PC6
## clinicalcategorical 0.818039 0.96659 3.387e-05 0.51898 0.9744 0.5411
## zymodemecategorical 0.001294 0.04593 3.079e-01 0.81772 0.1517 0.2662
## pathogenstrain      0.607150 0.07173 5.510e-06 0.43202 0.4455 0.7410
## passagenumber       0.007131 0.09458 3.026e-01 0.00987 0.3880 0.8164

test$cor_heatmap

sus_expt <- set_expt_conditions(lp_expt, fact = "sus_category") %>%
  set_expt_batches(fact = "zymodemecategorical")

sus_norm <- normalize_expt(sus_expt, transform = "log2", convert = "cpm",
                           norm = "quant", filter = TRUE)

## Removing 145 low-count genes (8565 remaining).

## transform_counts: Found 2 values equal to 0, adding 1 to the matrix.

sus_pca <- plot_pca(sus_norm, plot_title = "PCA of parasite expression values")
sus_pca$plot

sus_nb <- normalize_expt(sus_expt, transform = "log2", convert = "cpm",
                         batch = "svaseq", filter = TRUE)

## Removing 145 low-count genes (8565 remaining).

## batch_counts: Before batch/surrogate estimation, 568 entries are x==0: 0%.

## batch_counts: Before batch/surrogate estimation, 2107 entries are 0<x<1: 1%.

## Setting 110 low elements to zero.

## transform_counts: Found 110 values equal to 0, adding 1 to the matrix.

sus_nb_pca <- plot_pca(sus_nb, plot_title = "PCA of parasite expression values")
pp(file = "images/sus_nb_pca.png", image = sus_nb_pca$plot)

At this time, we do not have very many samples, so the set of metrics/plots is fairly limited. There is really only one factor in the metadata which we can use for performing differential expression analyses, the ‘zymodeme’.

5 Zymodeme analyses

The following sections perform a series of analyses which seek to elucidate differences between the zymodemes 2.2 and 2.3 either through differential expression or variant profiles.

5.1 Differential expression

5.1.1 With respect to zymodeme attribution

TODO: Do this with and without sva and compare the results.

zy_expt <- subset_expt(lp_expt, subset = "condition=='z2.2'|condition=='z2.3'")

## subset_expt(): There were 47, now there are 25 samples.

zy_norm <- normalize_expt(zy_expt, filter = TRUE, convert = "cpm", norm = "quant")

## Removing 167 low-count genes (8543 remaining).

zy_de_nobatch <- sm(all_pairwise(zy_expt, filter = TRUE, model_batch = "svaseq"))
zy_de <- sm(all_pairwise(zy_expt, filter = TRUE, model_batch = "svaseq"))
zy_table <- sm(combine_de_tables(zy_de, excel = glue::glue("excel/zy_tables-v{ver}.xlsx")))
zy_sig <- sm(extract_significant_genes(zy_table, excel = glue::glue("excel/zy_sig-v{ver}.xlsx")))

5.1.2 Images of zymodeme DE

zy_table[["plots"]][["z23_vs_z22"]][["deseq_ma_plots"]][["plot"]]

5.2 With respect to cure/failure

In contrast, we can search for genes which are differentially expressed with respect to cure/failure status.

cf_de <- sm(all_pairwise(cf_expt, filter = TRUE, model_batch = "svaseq"))

cf_table <- sm(combine_de_tables(cf_de, excel = glue::glue("excel/cf_tables-v{ver}.xlsx")))
cf_sig <- sm(extract_significant_genes(cf_table, excel = glue::glue("excel/cf_sig-v{ver}.xlsx")))

5.3 With respect to susceptibility

Finally, we can use our category of susceptibility and look for genes which change from sensitive to resistant. Keep in mind, though, that for the moment we have a lot of ambiguous and unknown strains.

sus_de <- sm(all_pairwise(sus_expt, filter = TRUE, model_batch = "svaseq"))

sus_table <- sm(combine_de_tables(sus_de, excel = glue::glue("excel/sus_tables-v{ver}.xlsx")))
sus_sig <- sm(extract_significant_genes(sus_table, excel = glue::glue("excel/sus_sig-v{ver}.xlsx")))

5.4 Ontology searches

Now let us look for ontology categories which are increased in the 2.3 samples followed by the 2.2 samples.

## Gene categories more represented in the 2.3 group.
zy_go_up <- sm(simple_goseq(sig_genes = zy_sig[["deseq"]][["ups"]][[1]],
                            go_db = lp_go, length_db = lp_lengths))

## Gene categories more represented in the 2.2 group.
zy_go_down <- sm(simple_goseq(sig_genes = zy_sig[["deseq"]][["downs"]][[1]],
                              go_db = lp_go, length_db = lp_lengths))

5.4.1 A couple plots from the differential expression

5.4.1.1 Number of genes in agreement among DE methods, 2.3 more than 2.2

In the function ‘combined_de_tables()’ above, one of the tasks performed is to look at the agreement among DESeq2, limma, and edgeR. The following show a couple of these for the set of genes observed with a fold-change >= |2| and adjusted p-value <= 0.05.

zy_table[["venns"]][[1]][["p_lfc1"]][["up_noweight"]]

5.4.1.2 Number of genes in agreement among DE methods, 2.2 more than 2.3

zy_table[["venns"]][[1]][["p_lfc1"]][["down_noweight"]]

5.4.1.3 MA plot of the differential expression between the zymodemes.

zy_table$plots[[1]][["deseq_ma_plots"]][["plot"]]

5.4.1.4 goseq ontology plots of groups of genes, 2.3 more than 2.2

zy_go_up$pvalue_plots$bpp_plot_over

5.4.1.5 goseq ontology plots of groups of genes, 2.2 more than 2.3

zy_go_down$pvalue_plots$bpp_plot_over

5.5 Zymodeme enzyme gene IDs

Najib read me an email listing off the gene names associated with the zymodeme classification. I took those names and cross referenced them against the Leishmania panamensis gene annotations and found the following:

They are:

1. ALAT: LPAL13_120010900 – alanine aminotransferase
2. ASAT: LPAL13_340013000 – aspartate aminotransferase
3. G6PD: LPAL13_000054100 – glucase-6-phosphate 1-dehydrogenase
4. NH: LPAL13_14006100, LPAL13_180018500 – inosine-guanine nucleoside hydrolase
5. MPI: LPAL13_320022300 (maybe) – mannose phosphate isomerase (I chose phosphomannose isomerase)

Given these 6 gene IDs (NH has two gene IDs associated with it), I can do some looking for specific differences among the various samples.

5.5.1 Expression levels of zymodeme genes

The following creates a colorspace (red to green) heatmap showing the observed expression of these genes in every sample.

my_genes <- c("LPAL13_120010900", "LPAL13_340013000", "LPAL13_000054100",
              "LPAL13_140006100", "LPAL13_180018500", "LPAL13_320022300",
              "other")
my_names <- c("ALAT", "ASAT", "G6PD", "NHv1", "NHv2", "MPI", "other")

zymo_expt <- exclude_genes_expt(zy_norm, ids = my_genes, method = "keep")

## Before removal, there were 8543 genes, now there are 6.

## There are 25 samples which kept less than 90 percent counts.

## TMRC20001 TMRC20005 TMRC20039 TMRC20037 TMRC20038 TMRC20041 TMRC20015 TMRC20009 
##    0.1310    0.1318    0.1299    0.1100    0.1128    0.1179    0.1146    0.1135 
## TMRC20010 TMRC20016 TMRC20011 TMRC20012 TMRC20013 TMRC20017 TMRC20014 TMRC20018 
##    0.1098    0.1059    0.1101    0.1205    0.1205    0.1063    0.1089    0.1144 
## TMRC20021 TMRC20022 TMRC20053 TMRC20052 TMRC20064 TMRC20051 TMRC20050 TMRC20062 
##    0.1061    0.1305    0.1182    0.1104    0.1138    0.1280    0.1151    0.1283 
## TMRC20054 
##    0.1276

zymo_heatmap <- plot_sample_heatmap(zymo_expt, row_label = my_names)
zymo_heatmap

5.6 Empirically observed Zymodeme genes from differential expression analysis

In contrast, the following plots take the set of genes which are shared among all differential expression methods (|lfc| >= 1.0 and adjp <= 0.05) and use them to make categories of genes which are increased in 2.3 or 2.2.

shared_zymo <- intersect_significant(zy_table)

## Deleting the file excel/intersect_significant.xlsx before writing the tables.

up_shared <- shared_zymo[["ups"]][[1]][["data"]][["all"]]
rownames(up_shared)

##  [1] "LPAL13_000033300" "LPAL13_000012000" "LPAL13_310031300" "LPAL13_000038400"
##  [5] "LPAL13_000038500" "LPAL13_000012100" "LPAL13_340039600" "LPAL13_050005000"
##  [9] "LPAL13_310031000" "LPAL13_310039200" "LPAL13_210015500" "LPAL13_350063000"
## [13] "LPAL13_140019300" "LPAL13_270034100" "LPAL13_340039700" "LPAL13_180013900"
## [17] "LPAL13_170015400" "LPAL13_350013200" "LPAL13_330021800" "LPAL13_140019100"
## [21] "LPAL13_240009700" "LPAL13_330021900" "LPAL13_140019200" "LPAL13_000052700"
## [25] "LPAL13_250025700" "LPAL13_350073200" "LPAL13_310028500" "LPAL13_320038700"
## [29] "LPAL13_210005000" "LPAL13_300031600" "LPAL13_110015700" "LPAL13_000045100"
## [33] "LPAL13_230011200" "LPAL13_040007800" "LPAL13_230011400" "LPAL13_290016200"
## [37] "LPAL13_310032500" "LPAL13_230011500" "LPAL13_140019400" "LPAL13_000010600"
## [41] "LPAL13_100005800"

upshared_expt <- exclude_genes_expt(zy_norm, ids = rownames(up_shared), method = "keep")

## Before removal, there were 8543 genes, now there are 41.

## There are 25 samples which kept less than 90 percent counts.

## TMRC20001 TMRC20005 TMRC20039 TMRC20037 TMRC20038 TMRC20041 TMRC20015 TMRC20009 
##    0.4245    0.1639    0.2311    0.6340    0.7125    0.2043    0.5332    0.1935 
## TMRC20010 TMRC20016 TMRC20011 TMRC20012 TMRC20013 TMRC20017 TMRC20014 TMRC20018 
##    0.4926    0.4036    0.2015    0.1613    0.4721    0.2591    0.2134    0.4470 
## TMRC20021 TMRC20022 TMRC20053 TMRC20052 TMRC20064 TMRC20051 TMRC20050 TMRC20062 
##    0.5088    0.1852    0.2650    0.5964    0.5949    0.8250    0.2773    0.8369 
## TMRC20054 
##    0.7197

We can plot a quick heatmap to get a sense of the differences observed between the genes which are different between the two zymodemes.

5.6.1 Heatmap of zymodeme gene expression increased in 2.3 vs. 2.2

high_23_heatmap <- plot_sample_heatmap(upshared_expt, row_label = rownames(up_shared))
high_23_heatmap

5.6.2 Heatmap of zymodeme gene expression increased in 2.2 vs. 2.3

down_shared <- shared_zymo[["downs"]][[1]][["data"]][["all"]]
downshared_expt <- exclude_genes_expt(zy_norm, ids = rownames(down_shared), method = "keep")

## Before removal, there were 8543 genes, now there are 63.

## There are 25 samples which kept less than 90 percent counts.

## TMRC20001 TMRC20005 TMRC20039 TMRC20037 TMRC20038 TMRC20041 TMRC20015 TMRC20009 
##    0.2181    0.6764    0.6475    0.1938    0.1849    0.6785    0.1786    0.6274 
## TMRC20010 TMRC20016 TMRC20011 TMRC20012 TMRC20013 TMRC20017 TMRC20014 TMRC20018 
##    0.1628    0.2041    0.5594    0.5529    0.1608    0.6469    0.6368    0.1568 
## TMRC20021 TMRC20022 TMRC20053 TMRC20052 TMRC20064 TMRC20051 TMRC20050 TMRC20062 
##    0.1565    0.6738    0.5544    0.1747    0.1908    0.1781    0.6052    0.1779 
## TMRC20054 
##    0.1921

high_22_heatmap <- plot_sample_heatmap(downshared_expt, row_label = rownames(down_shared))
high_22_heatmap

6 SNP profiles

Now I will combine our previous samples and our new samples in the hopes of finding variant positions which help elucidate currently unknown aspects of either group via their clustering to known samples from the other group. In other words, we do not know the zymodeme annotations for the old samples nor the strain identities (or the shortcut ‘chronic vs. self-healing’) for the new samples. I hope to make educated guesses given the variant profiles. There are some differences in how the previous and current data sets were analyzed (though I have since redone the old samples so it should be trivial to remove those differences now).

I added our 2016 data to a specific TMRC2 sample sheet, dated 20191203. Thus I will load the data here. That previous data was mapped using tophat, so I will also need to make some changes to the gene names to accomodate the two mappings.

old_expt <- sm(create_expt("sample_sheets/tmrc2_samples_20191203.xlsx",
                           file_column = "tophat2file"))

tt <- lp_expt$expressionset
rownames(tt) <- gsub(pattern = "^exon_", replacement = "", x = rownames(tt))
rownames(tt) <- gsub(pattern = "\\.E1$", replacement = "", x = rownames(tt))
lp_expt$expressionset <- tt

tt <- old_expt$expressionset
rownames(tt) <- gsub(pattern = "^exon_", replacement = "", x = rownames(tt))
rownames(tt) <- gsub(pattern = "\\.1$", replacement = "", x = rownames(tt))
old_expt$expressionset <- tt

6.1 Create the SNP expressionset

One other important caveat, we have a group of new samples which have not yet run through the variant search pipeline, so I need to remove them from consideration. Though it looks like they finished overnight…

## The next line drops the samples which are missing the SNP pipeline.
lp_snp <- subset_expt(lp_expt, subset="!is.na(pData(lp_expt)[['bcftable']])")

## subset_expt(): There were 47, now there are 46 samples.

new_snps <- sm(count_expt_snps(lp_snp, annot_column = "bcftable"))
old_snps <- sm(count_expt_snps(old_expt, annot_column = "bcftable", snp_column = 2))

both_snps <- combine_expts(new_snps, old_snps)
both_norm <- sm(normalize_expt(both_snps, transform = "log2", convert = "cpm", filter = TRUE))

## strains <- both_norm[["design"]][["strain"]]
both_norm <- set_expt_conditions(both_norm, fact = "strain")

The data structure ‘both_norm’ now contains our 2016 data along with the newer data collected since 2019.

6.2 Plot of SNP profiles for zymodemes

The following plot shows the SNP profiles of all samples (old and new) where the colors at the top show either the 2.2 strains (orange), 2.3 strains (green), the previous samples (purple), or the various lab strains (pink etc).

old_new_variant_heatmap <- plot_disheat(both_norm)
pp(file = "images/raw_snp_disheat.png", image = old_new_variant_heatmap,
   height = 12, width = 12)

The function get_snp_sets() takes the provided metadata factor (in this case ‘condition’) and looks for variants which are exclusive to each element in it. In this case, this is looking for differences between 2.2 and 2.3, as well as the set shared among them.

snp_sets <- get_snp_sets(both_snps, factor = "condition")

## The factor z2.3 has 14 rows.

## The factor z2.2 has 11 rows.

## The factor unknown has 12 rows.

## The factor null has 9 rows.

## The factor sh has 13 rows.

## The factor chr has 14 rows.

## The factor inf has 6 rows.

## Iterating over 727 elements.

both_expt <- combine_expts(lp_expt, old_expt)

snp_genes <- sm(snps_vs_genes(both_expt, snp_sets, expt_name_col = "chromosome"))
## I think we have some metrics here we can plot...
snp_subset <- sm(snp_subset_genes(
  both_expt, both_snps,
  genes = c("LPAL13_120010900", "LPAL13_340013000", "LPAL13_000054100",
            "LPAL13_140006100", "LPAL13_180018500", "LPAL13_320022300")))
## zymo_heat <- plot_sample_heatmap(snp_subset, row_label = rownames(exprs(snp_subset)))

Didn’t I create a set of densities by chromosome? Oh I think they come in from get_snp_sets()

6.3 SNPS associated with clinical response in the TMRC samples

clinical_sets <- get_snp_sets(new_snps, factor = "clinicalresponse")

## The factor cure has 17 rows.

## The factor failure has 15 rows.

## The factor laboratory line has only 1 row.

## The factor nd has 3 rows.

## The factor notapplicable has 7 rows.

## The factor reference strain has 3 rows.

## Iterating over 693 elements.

density_vec <- clinical_sets[["density"]]
chromosome_idx <- grep(pattern = "LpaL", x = names(density_vec))
density_df <- as.data.frame(density_vec[chromosome_idx])
density_df[["chr"]] <- rownames(density_df)
colnames(density_df) <- c("density_vec", "chr")
ggplot(density_df, aes_string(x = "chr", y = "density_vec")) +
  ggplot2::geom_col() +
  ggplot2::theme(axis.text = ggplot2::element_text(size = 10, colour = "black"),
                 axis.text.x = ggplot2::element_text(angle = 90, vjust = 0.5))

## clinical_written <- write_variants(new_snps)

6.3.1 Cross reference these variants by gene

clinical_genes <- sm(snps_vs_genes(lp_expt, clinical_sets, expt_name_col = "chromosome"))

snp_density <- merge(as.data.frame(clinical_genes[["summary_by_gene"]]),
                     as.data.frame(fData(lp_expt)),
                     by = "row.names")
snp_density <- snp_density[, c(1, 2, 4, 15)]
colnames(snp_density) <- c("name", "snps", "product", "length")
snp_density[["product"]] <- tolower(snp_density[["product"]])
snp_density[["length"]] <- as.numeric(snp_density[["length"]])
snp_density[["density"]] <- snp_density[["snps"]] / snp_density[["length"]]
snp_idx <- order(snp_density[["density"]], decreasing = TRUE)
snp_density <- snp_density[snp_idx, ]

removers <- c("amastin", "gp63", "leishmanolysin")
for (r in removers) {
  drop_idx <- grepl(pattern = r, x = snp_density[["product"]])
  snp_density <- snp_density[!drop_idx, ]
}
## Filter these for [A|a]mastin gp63 Leishmanolysin

clinical_snps <- snps_intersections(lp_expt, clinical_sets, chr_column = "chromosome")

head(as.data.frame(clinical_snps$inters[["Failure"]]))

## data frame with 0 columns and 0 rows

head(as.data.frame(clinical_snps$inters[["Cure"]]))

## data frame with 0 columns and 0 rows

head(clinical_snps$gene_summaries$Failure)

## NULL

head(clinical_snps$gene_summaries$Cure, n = 100)

## NULL

annot <- fData(lp_expt)
clinical_interest <- as.data.frame(clinical_snps[["gene_summaries"]][["Cure"]])
clinical_interest <- merge(clinical_interest, as.data.frame(clinical_snps[["gene_summaries"]][["Failure"]]), by = "row.names")
rownames(clinical_interest) <- clinical_interest[["Row.names"]]
clinical_interest[["Row.names"]] <- NULL
colnames(clinical_interest) <- c("cure_snps","fail_snps")

## Error in names(x) <- value: 'names' attribute [2] must be the same length as the vector [0]

annot <- merge(annot, clinical_interest, by = "row.names")
rownames(annot) <- annot[["Row.names"]]
annot[["Row.names"]] <- NULL
fData(lp_expt$expressionset) <- annot

7 Zymodeme for new samples

The heatmap produced here should show the variants only for the zymodeme genes.

7.1 Hunt for snp clusters

I am thinking that if we find clusters of locations which are variant, that might provide some PCR testing possibilities.

new_sets <- get_snp_sets(new_snps, factor = "phenotypiccharacteristics")

## The factor 22 has 11 rows.

## The factor 23 has 14 rows.

## The factor laboratory line has only 1 row.

## The factor notapplicable has 17 rows.

## The factor reference strain has 3 rows.

## Iterating over 693 elements.

summary(new_sets)

##               Length Class      Mode     
## medians         6    data.frame list     
## possibilities   5    -none-     character
## intersections  31    -none-     list     
## chr_data      693    -none-     list     
## set_names      32    -none-     list     
## invert_names   32    -none-     list     
## density       693    -none-     numeric

## 1000000: 2.2
## 0100000: 2.3

summary(new_sets[["intersections"]][["100000"]])

## Length  Class   Mode 
##      0   NULL   NULL

dim(new_sets$intersections[["100000"]])

## NULL

sequential_variants <- function(snp_sets, conditions = NULL, minimum = 3, maximum_separation = 3) {
  if (is.null(conditions)) {
    conditions <- 1
  }
  intersection_sets <- snp_sets[["intersections"]]
  intersection_names <- snp_sets[["set_names"]]
  chosen_intersection <- 1
  if (is.numeric(conditions)) {
    chosen_intersection <- conditions
  } else {
    intersection_idx <- intersection_names == conditions
    chosen_intersection <- names(intersection_names)[intersection_idx]
  }

  possible_positions <- intersection_sets[[chosen_intersection]]
  position_table <- data.frame(row.names = possible_positions)
  pat <- "^chr_(.+)_pos_(.+)_ref_.*$"
  position_table[["chr"]] <- gsub(pattern = pat, replacement = "\\1", x = rownames(position_table))
  position_table[["pos"]] <- as.numeric(gsub(pattern = pat, replacement = "\\2", x = rownames(position_table)))
  position_idx <- order(position_table[, "chr"], position_table[, "pos"])
  position_table <- position_table[position_idx, ]
  position_table[["dist"]] <- 0

  last_chr <- ""
  for (r in 1:nrow(position_table)) {
    this_chr <- position_table[r, "chr"]
    if (r == 1) {
      position_table[r, "dist"] <- position_table[r, "pos"]
      last_chr <- this_chr
      next
    }
    if (this_chr == last_chr) {
      position_table[r, "dist"] <- position_table[r, "pos"] - position_table[r - 1, "pos"]
    } else {
      position_table[r, "dist"] <- position_table[r, "pos"]
    }
    last_chr <- this_chr
  }

  sequentials <- position_table[["dist"]] <= maximum_separation

  ## The following can tell me how many runs of each length occurred, that is not quite what I want.
  ## Now use run length encoding to find the set of sequential sequentials!
  rle_result <- rle(sequentials)
  rle_values <- rle_result[["values"]]
  ## The following line is equivalent to just leaving values alone:
  ## true_values <- rle_result[["values"]] == TRUE
  rle_lengths <- rle_result[["lengths"]]
  true_sequentials <- rle_lengths[rle_values]
  rle_idx <- cumsum(rle_lengths)[which(rle_values)]

  position_table[["last_sequential"]] <- 0
  count <- 0
  for (r in rle_idx) {
    count <- count + 1
    position_table[r, "last_sequential"] <- true_sequentials[count]
  }

  wanted_idx <- position_table[["last_sequential"]] >= minimum
  wanted <- position_table[wanted_idx, c("chr", "pos")]
  return(wanted)
}

zymo22_sequentials <- sequential_variants(new_sets, conditions = "2.2")

## Error in intersection_sets[[chosen_intersection]]: attempt to select less than one element in get1index

zymo22_sequentials

## Error in eval(expr, envir, enclos): object 'zymo22_sequentials' not found

zymo23_sequentials <- sequential_variants(new_sets, conditions = "2.3")

## Error in intersection_sets[[chosen_intersection]]: attempt to select less than one element in get1index

zymo23_sequentials

## Error in eval(expr, envir, enclos): object 'zymo23_sequentials' not found

snp_genes <- sm(snps_vs_genes(lp_expt, new_sets, expt_name_col = "chromosome"))
new_zymo_norm  <- normalize_expt(new_snps, filter = TRUE, convert = "cpm", norm = "quant", transform = TRUE)

## Removing 0 low-count genes (558524 remaining).

## transform_counts: Found 11978651 values equal to 0, adding 1 to the matrix.

new_zymo_norm <- set_expt_conditions(new_zymo_norm, fact = "phenotypiccharacteristics")
zymo_heat <- plot_disheat(new_zymo_norm)

zymo_subset <- snp_subset_genes(lp_expt, new_snps,
                                genes = c("LPAL13_120010900", "LPAL13_340013000", "LPAL13_000054100",
                                        "LPAL13_140006100", "LPAL13_180018500", "LPAL13_320022300"))

## Error: cannot subset by character when names are NULL

zymo_subset <- set_expt_conditions(zymo_subset, fact = "phenotypiccharacteristics")

## Error in h(simpleError(msg, call)): error in evaluating the argument 'object' in selecting a method for function 'pData': object 'zymo_subset' not found

## zymo_heat <- plot_sample_heatmap(zymo_subset, row_label = rownames(exprs(snp_subset)))

des <- both_norm$design
undef_idx <- is.na(des[["strain"]])
des[undef_idx, "strain"] <- "unknown"

##hmcols <- colorRampPalette(c("yellow","black","darkblue"))(256)
correlations <- hpgl_cor(exprs(both_norm))

zymo_missing_idx <- is.na(des[["phenotypiccharacteristics"]])
des[zymo_missing_idx, "phenotypiccharacteristics"] <- "unknown"

## Warning in `[<-.factor`(`*tmp*`, iseq, value = c("unknown", "unknown",
## "unknown", : invalid factor level, NA generated

mydendro <- list(
  "clustfun" = hclust,
  "lwd" = 2.0)
col_data <- as.data.frame(des[, c("phenotypiccharacteristics", "clinicalcategorical")])
unknown_clinical <- is.na(col_data[["clinicalcategorical"]])
row_data <- as.data.frame(des[, c("strain")])
colnames(col_data) <- c("zymodeme", "outcome")
col_data[unknown_clinical, "outcome"] <- "undefined"

## Warning in `[<-.factor`(`*tmp*`, iseq, value = c("undefined", "undefined", :
## invalid factor level, NA generated

colnames(row_data) <- c("strain")
myannot <- list(
  "Col" = list("data" = col_data),
  "Row" = list("data" = row_data))
myclust <- list("cuth" = 1.0,
                "col" = BrewerClusterCol)
mylabs <- list(
  "Row" = list("nrow" = 4),
  "Col" = list("nrow" = 4))
hmcols <- colorRampPalette(c("darkblue", "beige"))(240)
map1 <- annHeatmap2(
  correlations,
  dendrogram = mydendro,
  annotation = myannot,
  cluster = myclust,
  labels = mylabs,
  ## The following controls if the picture is symmetric
  scale = "none",
  col = hmcols)

## Warning in breakColors(breaks, col): more colors than classes: ignoring 29 last
## colors

pp(file = "images/dendro_heatmap.png", image = map1, height=12, width = 12)

## annotated Heatmap
## 
## Rows: 'dendrogram' with 2 branches and 79 members total, at height 5.258 
##   11  annotation variable(s)
## Cols: 'dendrogram' with 2 branches and 79 members total, at height 5.258 
##   9  annotation variable(s)

## plot(map1)

8 Using Variant profiles to make guesses about strains and chronic/self-healing

The following uses the same information to make some guesses about the strains used in the new samples.

des <- both_norm$design
undef_idx <- is.na(des[["strain"]])
des[undef_idx, "strain"] <- "unknown"
##hmcols <- colorRampPalette(c("yellow","black","darkblue"))(256)
correlations <- hpgl_cor(exprs(both_norm))

mydendro <- list(
  "clustfun" = hclust,
  "lwd" = 2.0)
col_data <- as.data.frame(des[, c("condition")])
row_data <- as.data.frame(des[, c("strain")])
colnames(col_data) <- c("condition")
colnames(row_data) <- c("strain")
myannot <- list(
  "Col" = list("data" = col_data),
  "Row" = list("data" = row_data))
myclust <- list("cuth" = 1.0,
                "col" = BrewerClusterCol)
mylabs <- list(
  "Row" = list("nrow" = 4),
  "Col" = list("nrow" = 4))
hmcols <- colorRampPalette(c("darkblue", "beige"))(170)
map1 <- annHeatmap2(
  correlations,
  dendrogram = mydendro,
  annotation = myannot,
  cluster = myclust,
  labels = mylabs)
##  col = hmcols)
plot(map1)

pheno <- subset_expt(lp_expt, subset = "condition=='z2.2'|condition=='z2.3'")

## subset_expt(): There were 47, now there are 25 samples.

pheno <- subset_expt(pheno, subset="!is.na(pData(pheno)[['bcftable']])")

## subset_expt(): There were 25, now there are 25 samples.

pheno_snps <- sm(count_expt_snps(pheno, annot_column = "bcftable"))

xref_prop <- table(pheno_snps$conditions)
pheno_snps$conditions

##  [1] "z2.3" "z2.2" "z2.2" "z2.3" "z2.3" "z2.2" "z2.3" "z2.2" "z2.3" "z2.3"
## [11] "z2.2" "z2.2" "z2.3" "z2.2" "z2.2" "z2.3" "z2.3" "z2.2" "z2.2" "z2.3"
## [21] "z2.3" "z2.3" "z2.2" "z2.3" "z2.3"

idx_tbl <- exprs(pheno_snps) > 5
new_tbl <- data.frame(row.names = rownames(exprs(pheno_snps)))
for (n in names(xref_prop)) {
  new_tbl[[n]] <- 0
  idx_cols <- which(pheno_snps[["conditions"]] == n)
  prop_col <- rowSums(idx_tbl[, idx_cols]) / xref_prop[n]
  new_tbl[n] <- prop_col
}
new_tbl[["ratio"]] <- (new_tbl[["z2.2"]] - new_tbl[["z2.3"]])
keepers <- grepl(x = rownames(new_tbl), pattern = "LpaL13")
new_tbl <- new_tbl[keepers, ]
new_tbl[["SNP"]] <- rownames(new_tbl)
new_tbl[["Chromosome"]] <- gsub(x = new_tbl[["SNP"]], pattern = "chr_(.*)_pos_.*", replacement = "\\1")
new_tbl[["Position"]] <- gsub(x = new_tbl[["SNP"]], pattern = ".*_pos_(\\d+)_.*", replacement = "\\1")
new_tbl <- new_tbl[, c("SNP", "Chromosome", "Position", "ratio")]
library(CMplot)

## Much appreciate for using CMplot.

## Full description, Bug report, Suggestion and the latest codes:

## https://github.com/YinLiLin/CMplot

CMplot(new_tbl)

##  SNP-Density Plotting.

##  Circular-Manhattan Plotting ratio.
##  Rectangular-Manhattan Plotting ratio.

##  QQ Plotting ratio.

##  Plots are stored in: /mnt/cbcb/fs01_abelew/cbcb-lab/nelsayed/scratch/atb/rnaseq/lpanamensis_tmrc_2019

if (!isTRUE(get0("skip_load"))) {
  pander::pander(sessionInfo())
  message(paste0("This is hpgltools commit: ", get_git_commit()))
  message(paste0("Saving to ", savefile))
  tmp <- sm(saveme(filename = savefile))
}

## If you wish to reproduce this exact build of hpgltools, invoke the following:

## > git clone http://github.com/abelew/hpgltools.git

## > git reset 72947fcc6afe09da22d71967059edd84e3063341

## This is hpgltools commit: Tue Jun 1 15:57:56 2021 -0400: 72947fcc6afe09da22d71967059edd84e3063341

## Saving to tmrc2_02sample_estimation_v202106.rda.xz

tmp <- loadme(filename = savefile)