cluster_loading.py

gistfile1.txt

analysis_dir = "/nas3/gpratt/projects/public_clip/analysis/v5"

experiment_mapping = {
                      "peaks" : "clipper",
                      "ripseeker" : "ripseeker",
                      "pyicoclip" : "pyicoclip",
                      "piranha" : "piranha",
                      "bed_notrim_ingenes_clusters_hg1950" : "kasey",
                      "bed_notrim_ingenes_clusters_mm950"  : "kasey",
                      }

peak_files = glob.glob(os.path.join(analysis_dir, "*.bed"))
samples = itertools.groupby(sorted(peak_files), lambda x: os.path.basename(x).split(".")[0])

sample_dict = defaultdict(dict)
for sample, experiments in samples:
    for experiment in experiments:
        experiment_type = os.path.basename(experiment).split(".")[-2]
        sample_dict[sample][experiment_mapping[experiment_type]] = experiment
        
sample_df = pd.DataFrame(sample_dict)