ladder map

Matches ladders to peaks by correlation for fragment analysis. The strategy resembles the one used by Fragman for R.

One difference is that combinations of peaks are generated using NetworkX to eliminate impossible combinations. This reduces complexity substantially and allows for an exhaustive search to identify the best match.

Install

pip install peak-a-py

conda install -c bioconda peak-a-py

Usage

The library exists of two main classes, namely LadderMap and PeakArea. LadderMap matches ladders to peaks by correlation and stores peak and ladder information. The have methods to reassign timeseries data to basepair steps using linear regression.

PeakArea calculates peak area.

Example Usage

Area under curve using naive integrals

from fragment_analyzer.ladder_map import LadderMap, PeakArea

data = "demo/4071_Dx 230113_PRT1_PRT3_rn/PRT3_NA18507_4071_D11_Dx.fsa"

laddermap = LadderMap(data)

peak_area = PeakArea(
    laddermap.adjusted_step_dataframe(),
    start=190, 
    end=200,
    rel_height=.97
)

peak_area.plot_peak_widths()

Output

Visualization of best sample ladder peaks

fig = laddermap.plot_best_sample_ladder()

Output

Fitting model to the data

Voigt Distribution

peak_area.plot_lmfit_model("voigt")

Output

Gauss Distribution

peak_area.plot_lmfit_model("gauss")

Output

Looking at more than two peaks

peak_area = PeakArea(
    laddermap.adjusted_step_dataframe(channel="DATA1"),
    start=250, 
    end=300,
    num_peaks=4,
    padding=2,
    model="voigt"
)


peak_area.plot_lmfit_model()

The last peak is divided by the mean of the peaks to the left of it:

Output

If data needs baseline correction and normalization:

laddermap = LadderMap(data, normalize_peaks=False)

Messy output

Output

Normalized data:

laddermap = LadderMap(data, normalize_peaks=True)

Normalized peaks:

Output

Generate report:

from fragment_analyzer import LadderMap, PeakArea, generate_report

data = "demo/4062_Dx/3_PRT_2_4062_C02_Dx.fsa"
laddermap = LadderMap(data, normalize_peaks=False)
peak_area = PeakArea(
    laddermap.adjusted_step_dataframe(channel="DATA1"),
    start=200, 
    end=250,
    num_peaks=2,
    padding=2,
    model="gauss"
)

generate_report(laddermap, peak_area, name="my_folder/my-report")

The report is saves in my_folder as my-report.html. An example report can be found in examples

TODO

• output excel or csv with peak area, position of peak and height
• make agnostic algorithm of how many peaks one expects