Gene amplification was assessed by a FISH ratio (HER2 gene signals to chromosome 17 signals) of >2.2.
We further normalized for signal intensity among different experiments using averages of all the gene signals obtained during individual experiments.
We therefore chose probes containing LNA residues in suspension arrays to elevate discrimination and maintain high gene signals.
ATM deficiencies increased the incidence of cells with the split MLL gene signals in FISH analysis after etoposide treatment (Figure 1C).
Also, wavelet coefficients for gene signals from different datasets were compared at different scales (by computing RSS -Residual Sum of Squares- values), in order to assess which normalisation techniques bring the data closer together.
A significant increase in cells carrying split MLL gene signals at 6, 36 and 48 hours after etoposide exposure was observed when the cells were ATM deficient and not when they were ATM proficient (Figures 1B, C and S2).
Co-localisation was assessed by identifying on cells of ICF Patient 1 and Control 1, and ICF Patient 2 and Control 2, gene signals that showed any degree of overlap with the satellite DNA signal.
This shows that for levels 1,2 and 3, (corresponding to higher frequencies), PMOnly methods show the largest differences between gene signals, for most genes analysed, while LoessOnly and PMLoess are comparable.
Gene signals were normalized between experiments by quantile normalization [ 17].
Second, the gene signals needed to exceed 100.
The correlation of low RNA derived gene signals to gene signals obtained from standard RNA was poor for less to moderately abundant genes.
