HuProt CSF Analysis

Preprocessing

Read ../rawdata_huprot/Control_1256587_2000143978.gpr 
Read ../rawdata_huprot/Control_1311449_2000143966.gpr 
Read ../rawdata_huprot/Control_1312497_2000143964.gpr 
Read ../rawdata_huprot/Control_1313058_2000143976.gpr 
Read ../rawdata_huprot/Grade_L_CJ_14434_2000143958.gpr 
Read ../rawdata_huprot/Grade_L_CH_27096_2000155729.gpr 
Read ../rawdata_huprot/Grade_L_CJ_3659_2000143955.gpr 
Read ../rawdata_huprot/Grade_III_CH_15796_2000143692.gpr 
Read ../rawdata_huprot/Grade_III_CJ_11032_2000143957.gpr 
Read ../rawdata_huprot/Grade_III_CJ_27572_2000144442.gpr 
Read ../rawdata_huprot/Grade_III_CJ_9761_2000144443.gpr 
Read ../rawdata_huprot/Grade_III_CK_5121_2000144441.gpr 
Read ../rawdata_huprot/Grade_III_CK_6451_2000144440.gpr 
Read ../rawdata_huprot/Grade_IV_CH_12724_2000143977.gpr 
Read ../rawdata_huprot/Grade_IV_CH_29738_2000143975.gpr 
Read ../rawdata_huprot/Grade_IV_CH_31148_2000143961.gpr 
Read ../rawdata_huprot/Grade_IV_CH_32705_2000155727.gpr 
Read ../rawdata_huprot/Grade_IV_CJ_11291_2000143969.gpr 
Read ../rawdata_huprot/Grade_IV_CJ_12441_2000143979.gpr 
Read ../rawdata_huprot/Grade_IV_CK_6570_2000143967.gpr 

Intensity distribution pre normalization

Intensity distribution post normalization

PCA pre normalization

Performing principle component analysis on raw intensities, which have not been background corrected yields no separation for Low or high grade samples from the controls.

PCA post normalization

MDS post normalization

Heatmap post normalization

Differential Expression

Since the sample size is too less, we do not remove any samples even though they look like outliers on the MDS plot.

Similarity with Protoarray results

In terms of total common genes between the two assays, we obtain the following table. It is meant to be presented as a sanity check.

targets.protoarray <- readTargets('../rawdata/Protoarray_annotation.csv', sep=',')

targets.protoarray

targets

Conclusion

The table above represents the number of DE genes (adjusted p-value < 0.01) for protoarray and huprot assays. The ‘common’ column represents the number of common genes. ‘jaccard’ is a similarity metric.

We do not see ANY overlap between the two assays. There are two chief reasons:

1. The experiment is not balanaced: there are different number of samples involved at each level: control, grade2, grade3, grade4.

2. Sample size is too small: As evident from the PCA and MDS plots, there is too much heterogenity intra-grade and among controls.

Ideally point #2 should not affect the similarity we expect between huprot and protoarray results, but probably problems arising from #1 overshadow it.

At present, the results make little sense since there is no overlap at all.