plot.py

import pandas as pd
import numpy as np
import networkx as nx
import matplotlib.pyplot as plt
import matplotlib.cm as cm
import matplotlib.colors as mcolors
import seaborn as sns
import plotly.graph_objects as go
from adjustText import adjust_text
from gprofiler import GProfiler
from sankeyflow import Sankey


def plot_cci(graph, colors, plt_name, coords, pg, emax=None, leg=False, low=25, high=75, ignore_alpha=False, log=False, efactor=8, vfactor=12, vnames=True, figsize=None, scale_factor=2, node_size=2, font_size=10,return_figure=False):
    """
    This function does a CCI plot

    Parameters
    ----------
    graph :
        Paths of single condition LR data
    colors :
        Cell type (Cluster) Colors
    plt_name :
        Plot Name (Title)
    coords :
        object coordinates
    emax :
        Max MeanLR across the all inputs, if its not defined, the method going to consider the max find within a sample
    leg :
        Set color legend
    low :
        Lower threshold: This parameter low and high defines the edges
    high :
        Higher threshould which will be filtered. Edges within the interval [low\,high] are filtered.
    ignore_alpha :
        Not include transparency on the plot edges
    log :
        Logscale the interactions
    efactor :
        Edge scale factor
    vfactor :
        Certex scale factor
    vnames :
        Remove vertex labels
    pg :
        Pagerank values
    figsize:
        Set matplotlib figsize
    return_figure:
        Option for return matplotlib figure axes

    Returns
    -------
    Python default plot
    
    """
    graph = nx.from_pandas_edgelist(graph,
                                    source='source',
                                    target='target',
                                    edge_attr=True,
                                    create_using=nx.DiGraph())

    # Check Maximal Weight
    if emax is None:
        emax = 0
        for _, _, d in graph.edges(data=True):
            weight = d.get('weight', 0)
            if isinstance(weight, (int, float)):
                emax = max(emax, abs(weight))

    # Create colormap
    colors_list = sns.color_palette("coolwarm", 201)  # Adjust to match the R colormap
    col_pallet_colors = [colors_list[i] for i in range(201)]
    col_pallet_colors[10] = '#B8b9ba'  # Expand the palette range

    # Scale coordinates
    coords_array = np.array(list(coords.values()))

    if coords_array.shape[0] != 1:
        coords_mean = (np.mean(coords_array[:, 0]), np.mean(coords_array[:, 1]))
        coords_std = (np.std(coords_array[:, 0]), np.std(coords_array[:, 1]))
        coords_scale = {node: tuple((coord - mean) / std for coord, mean, std in zip(coords[node], coords_mean, coords_std)) for node in coords}
    else:
        coords_scale = coords

    coords_scale = {key: (coords_scale[key][0] * scale_factor, coords_scale[key][1] * scale_factor) for key in coords_scale}

    # Calculate edge colors and alpha
    edge_colors = []
    alpha = []

    for u, v, d in graph.edges(data=True):
        weight = d.get('weight', 0)
        we = np.round(np.interp(weight, [-emax, emax], [1, 200]))
        edge_colors.append(col_pallet_colors[int(we)])
        alpha_cond = low < d.get('inter', 0) < high and not np.isnan(d.get('inter', 0) < high)
        alpha.append(0 if alpha_cond else d.get('inter', 0) < high)

    # Set edge attributes
    for u, v, d in graph.edges(data=True):
        d['color'] = [(c[0], c[1], c[2], a) for c, a in zip(edge_colors, alpha)]
        if log:
            d['width'] = np.log2(1 + d.get('inter', 0)) * efactor if d.get('inter', 0) != 0 else 0
        else:
            d['width'] = d.get('inter', 0) * efactor if d.get('inter', 0) != 0 else 0
        d['arrow_size'] = 0.4
        d['arrow_width'] = d['width'] + 0.8
        d['loop_angle'] = np.nan

    node_colors = [str(colors.get(node)) for node in graph.nodes()]
    node_sizes = [size*1000*node_size for size in pg]

    # Plot the graph
    if figsize is None:
        figsize = (6,6)
    fig, ax = plt.subplots(figsize=figsize)
    
    nx.draw(graph, pos=coords_scale, edge_color=edge_colors, node_color=node_colors, node_size=node_sizes,
            width=[d['width'] for _, _, d in graph.edges(data=True)],
            arrows=True, arrowsize=30, arrowstyle='-|>',
            connectionstyle='arc3,rad=0.3', with_labels=vnames, font_size=font_size)
    ax.set_xlim(-4, 4)
    ax.set_ylim(-4, 4)

    # Node Pagerank legend
    if pg is not None:
        min_pg, max_pg = min(pg), max(pg)
        legend1 = ax.legend(loc='lower left', title="Pagerank",
                handles=[plt.Line2D([], [], linestyle='', marker='o', markersize=v / vfactor, markerfacecolor='black', markeredgecolor='none') for v in [min_pg, (min_pg + max_pg) / 2, max_pg]],
                labels=[round(min_pg, 2), round((min_pg + max_pg) / 2, 2), round(max_pg, 2)], bbox_to_anchor=(0.95, 0.3))

    # Thickness legend
    non_zero_inter_edges = [d['inter'] for _, _, d in graph.edges(data=True) if d.get('inter', 0) != 0]
    if non_zero_inter_edges:
        e_wid_sp = [round(min(non_zero_inter_edges), 2), round(min(non_zero_inter_edges) + (emax / 2), 2), round(emax, 2)]
        legend2 = ax.legend(e_wid_sp, title='Percentage of \nthe interactions', title_fontsize='small', loc='upper left', bbox_to_anchor=(0.95, 0.7))

    ax.add_artist(legend1)
    ax.set_title(plt_name)
    # Show the plot
    plt.tight_layout()
    plt.show()
    if return_figure:
        return (fig,ax)


def plot_pca_LR_comparative(lrobj_tblPCA, pca_table, dims=(1, 2), ret=False, ggi=True, include_tf=False, gene_types="all"):
    """
    This function is a proxy to the PCA plot in comparative conditions

    Parameters
    ----------
    lrobj_tblPCA :
        LRobject table with all data
    pca_table :
        table entry
    dims :
        PCA dims
    ret :
        return plot
    ggi :
        GGI mode
    include_tf :
        intracellular option
    gene_types :
        filter option of genes
    
    Returns
    -------
    Python default plot
    
    """
    
    pca_plot = {}
    # Extract PCA results and create a DataFrame
    pca_result = lrobj_tblPCA['pca'][pca_table]
    pca_df = lrobj_tblPCA['rankings'][pca_table]
    pca_df[['PC1', 'PC2']] = pca_result # Adjust dims to be zero-indexed
    pca_df = pca_df.set_index("nodes")

    if ggi:
        # Filter for LR or TF
        if gene_types == "LR":
            result_split_names = [name for name in pca_df.index if "|R" in name or "|L" in name]
        elif gene_types == "TF":
            result_split_names = [name for name in pca_df.index if "|TF" in name]
        else:
            result_split_names = pca_df.index.tolist()

        pca_df = pca_df.loc[result_split_names]

        # Mapping Table
        if include_tf:
            map_df = pd.DataFrame({"gene": pca_df.index})
            map_df["mapping"] = map_df["gene"].apply(lambda gene: "Receptor" if "|R" in gene else ("Ligand" if "|L" in gene else "Transcription Factor"))
            color_groups = ["#f8756b", "#00b835", "#619cff"]
        else:
            l_mapping = lrobj_tblPCA["tables"][pca_table.replace('_ggi', '')][["ligpair", "type_gene_A"]].rename(columns={"ligpair": "gene", "type_gene_A": "mapping"}).drop_duplicates()
            r_mapping = lrobj_tblPCA["tables"][pca_table.replace('_ggi', '')][["recpair", "type_gene_B"]].rename(columns={"recpair": "gene", "type_gene_B": "mapping"}).drop_duplicates()
            map_df = pd.concat([l_mapping, r_mapping]).drop_duplicates().reset_index(drop=True)
            map_df = map_df[map_df["gene"].isin(pca_df.index)]
            color_groups = ["#f8756b", "#00b835"]

        # Merge mapping info with PCA data
        pca_df = pca_df.merge(map_df, left_index=True, right_on="gene")

        #Threshold

        sdev_x = pca_df['PC1'].std()
        sdev_y = pca_df['PC2'].std()
        ver_zx = np.abs(pca_df['PC1']) >= (4 * sdev_x)
        ver_zy = np.abs(pca_df['PC2']) >= (4 * sdev_y)

        # Plotting
        x_max = max(abs(pca_df['PC1']))
        y_max = max(abs(pca_df['PC2']))

        pca_df['PC1'] = -pca_df['PC1']
        plt.figure(figsize=(10, 7))
        sns.scatterplot(x='PC1', y='PC2', data=pca_df, s=20, hue='mapping', palette=color_groups)

        # Adjust text labels to avoid overlap
        texts = []
        for i, gene in enumerate(pca_df['gene']):
            if ver_zx.iloc[i] or ver_zy.iloc[i]: 
                texts.append(plt.text(pca_df.loc[pca_df['gene'] == gene, 'PC1'].values[0], pca_df.loc[pca_df['gene'] == gene, 'PC2'].values[0], gene, fontsize=8,  bbox=dict(facecolor='white', edgecolor='black', boxstyle='round,pad=0.3')))
        
        adjust_text(texts,
                    force_text=(1.0, 2.0))

        plt.xlim(-x_max, x_max)
        plt.ylim(-y_max, y_max)
        plt.xlabel(f'PC{dims[0]}')
        plt.ylabel(f'PC{dims[1]}')
        plt.title(pca_table, y=1.08)
        plt.legend(title='Gene Type')
        plt.grid(True)

        plt.axhline(0, linestyle='--', color='gray')
        plt.axvline(0, linestyle='--', color='gray')
        
        plt.show()

        pca_plot[pca_table] = plt
        
    else:
        # No GGI
        x_max = max(abs(pca_df['PC1']))
        y_max = max(abs(pca_df['PC2']))

        pca_df['PC1'] = -pca_df['PC1']
        plt.figure(figsize=(10, 7))
        sns.scatterplot(x='PC1', y='PC2', data=pca_df)

        # Adjust text labels to avoid overlap
        texts = []
        for i, gene in enumerate(pca_df.index):
            texts.append(plt.text(pca_df.loc[gene, 'PC1'], pca_df.loc[gene, 'PC2'], gene, fontsize=8))
        
        adjust_text(texts, arrowprops=dict(arrowstyle='->', color='red'))

        plt.xlim(-x_max, x_max)
        plt.ylim(-y_max, y_max)
        plt.xlabel(f'PC{dims[0]}')
        plt.ylabel(f'PC{dims[1]}')
        plt.title(pca_table)
        plt.grid(True)

        # Set x and y axis intervals
        plt.xticks(np.arange(-np.ceil(x_max), np.ceil(x_max) + 1))
        plt.yticks(np.arange(-np.ceil(y_max), np.ceil(y_max) + 1))

        plt.axhline(0, linestyle='--', color='gray')
        plt.axvline(0, linestyle='--', color='gray')
        plt.show()

        pca_plot[pca_table] = plt

    if ret:
        return pca_plot
    

def plot_bar_rankings(annData, table_name, ranking, type = None, filter_sign = None, mode = "cci", top_num = 10):
    """
    This function generates the barplot for a given network ranking on the CGI level. Further, the genes can be filtered by selected gene types to filter the plot.

    Parameters
    ----------
    annData :
        AnnData object with all data

    table_name :
        name of the ranking table

    ranking :
        name of the network ranking to use

    type :
        gene type (L,R,TF, LR/RL, RTF/TFR, LTF/TFL)

    filter_sign :
        show all (NULL), only positive (pos), or only negativ (neg) results
    
    Returns
    -------
    Python default plot

    """

    if '_x_' in table_name:
        rankings_table = annData.uns['pycrosstalker']['results']['rankings'][table_name]

        if type is not None:
            if len(type) == 1:
                rankings_table = rankings_table[rankings_table['nodes'].str.contains(r'\|' + type)]
            elif len(type) == 2:
                if type == 'TF':
                    rankings_table = rankings_table[rankings_table['nodes'].str.contains(r'\|' + type)]
                else:
                    rankings_table = rankings_table[rankings_table['nodes'].str.contains(r'\|R|\|L')]
            elif len(type) == 3:
                if type in ['RTF', 'TFR']:
                    rankings_table = rankings_table[rankings_table['nodes'].str.contains(r'\|R|\|TF')]
                elif type in ['LTF', 'TFL']:
                    rankings_table = rankings_table[rankings_table['nodes'].str.contains(r'\|L|\|TF')]

        rankings_table = rankings_table.sort_values(by=ranking)
        
        if mode == 'cgi':
            rankings_table = pd.concat([rankings_table.head(top_num), rankings_table.tail(top_num)])
            # rankings_table.loc[rankings_table[ranking].abs().nlargest(20).index]
        else:
            pass

        if filter_sign == 'pos':
            rankings_table = rankings_table[rankings_table[ranking] > 0]
        elif filter_sign == 'neg':
            rankings_table = rankings_table[rankings_table[ranking] < 0]


        if rankings_table.empty:
            return "No entries with provided Filters."

        rankings_table['signal'] = ['negative' if x < 0 else 'positive' for x in rankings_table[ranking]]

        custom_palette = {'positive': '#FF6E00', 'negative':'#00FFFF'}  # Orange and Blue

        # Plot
        plt.figure(figsize=(8, 6))
        sns.barplot(x=ranking, y='nodes', data=rankings_table, hue='signal', dodge=False, palette=custom_palette)
        plt.title(f"Ranking for {table_name}")
        plt.xlabel(ranking)
        plt.ylabel('Nodes')

        # Set x-axis tick intervals
        max_val = rankings_table[ranking].max()
        min_val = rankings_table[ranking].min()
        ticks = np.linspace(min_val, max_val, num=5)  # Adjust 'num' for more/less intervals
        plt.xticks(ticks, [f'{tick:.2f}' for tick in ticks])

        # Invert y-axis to have highest values at the top
        plt.gca().invert_yaxis()

        # Show the legend only once
        handles, labels = plt.gca().get_legend_handles_labels()
        plt.legend(handles, labels, loc='lower right')

        plt.grid(True, linestyle='--', linewidth=0.5)
        plt.gca().set_axisbelow(True)
        # Adjust layout and show plot
        plt.tight_layout()
        plt.show()


def plot_sankey(lrobj_tbl, target = None, ligand_cluster = None, receptor_cluster = None, plt_name = None, threshold = 50, tfflag = True):
    """
    This function selected genes sankey plot

     Parameters
    ----------
    lrobj_tbl :
        LRobject table with all data

    target :
        gene

    ligand_cluster :
        Ligand Clusters

    receptor_cluster :
        Receptor Clusters

    plt_name :
        plot title

    threshold :
        top_n n value
    
    Returns
    -------
    Python default plot

    """

    lrobj_tbl = lrobj_tbl[(lrobj_tbl['type_gene_A'] == "Ligand") & (lrobj_tbl['type_gene_B'] == "Receptor")]

    if target is not None:
        if len(target.split('|')) > 1:
            target_type = str(target.split('|')[1])
            if target_type == 'R':
                if lrobj_tbl['gene_B'].str.contains('\\|').any():
                    pass
                else:
                    target = target.split('|')[0]
                data = lrobj_tbl[lrobj_tbl['gene_B'] == target]
            elif target_type == 'L':
                if lrobj_tbl['gene_A'].str.contains('\\|').any():
                    pass
                else:
                    target = target.split('|')[0]
                data = lrobj_tbl[lrobj_tbl['gene_A'] == target]
        else:
            data = lrobj_tbl[lrobj_tbl['allpair'].str.contains(target)]
    else:
        data = lrobj_tbl

    
    if ligand_cluster is not None:
        data = data[data['source'].isin(ligand_cluster)]
    
    if receptor_cluster is not None:
        data = data[data['target'].isin(receptor_cluster)]

    color_palette = ['#00BFC4', '#FF3E3E']

    
    if len(data) >= 1:
        cat_cols = ['source', 'gene_A', 'gene_B', 'target']
        value_cols = 'LRScore'
        data = data.loc[data['LRScore'].abs().nlargest(min(len(data), threshold)).index]
        title = plt_name

        gen_sankey(data, cat_cols, value_cols, title)
    
    else:
        print(f"Gene->{target} Not Found")
    

def gen_sankey2(df, cat_cols=[], value_cols='', title='Sankey Diagram'):
    """
    Helper function to the function plot_sankey()

     Parameters
    ----------
    df :
        Dataframe

    cat_cols :
        Columns interested in the sankey plot

    value_cols :
        Sankey plot generated using connections based on this value_cols

    title :
        Title of Sankey plot
    
    Returns
    -------
    Nothing (plots Sankey plot)

    """

    df['source'] += 'S'
    df['target'] += 'T'
    
    labelList = []
    for catCol in cat_cols:
        labelListTemp =  list((df[catCol].values))
        labelList = labelList + labelListTemp    
        
    for i in range(len(cat_cols)-1):
        if i==0:
            sourceTargetDf = df[[cat_cols[i],cat_cols[i+1],value_cols]]
            sourceTargetDf.columns = ['source','target','count']
        else:
            tempDf = df[[cat_cols[i],cat_cols[i+1],value_cols]]
            tempDf.columns = ['source','target','count']
            sourceTargetDf = pd.concat([sourceTargetDf,tempDf])
        # sourceTargetDf = sourceTargetDf.groupby(['source','target']).agg({'count':'sum'}).reset_index()
        
    sourceTargetDf['sourceID'] = sourceTargetDf['source'].apply(lambda x: labelList.index(x))
    sourceTargetDf['targetID'] = sourceTargetDf['target'].apply(lambda x: labelList.index(x))

    for i, label in enumerate(labelList):
        if label[-1:] == 'S' or label[-1:] == 'T':
            labelList[i] = labelList[i][:-1]
        
    norm = mcolors.Normalize(vmin=min(sourceTargetDf['count']), vmax=max(sourceTargetDf['count']))
    colormap = cm.get_cmap('RdBu_r')
    link_colors = [mcolors.to_hex(colormap(norm(value))) for value in sourceTargetDf['count']]
    
    fig =  go.Figure(data = [go.Sankey(
        node = dict(
          pad = 0,
          thickness = 20,
          line = dict(
            color = "black",
            width = 0.5
          ),
          label = labelList,
          color = "white"
        ),
        link = dict(
          source = sourceTargetDf['sourceID'],
          target = sourceTargetDf['targetID'],
          value = [abs(i) for i in sourceTargetDf['count']],
          color = link_colors
        )    
    )])
    
    colorbar_trace = go.Scatter(
        x=[None], y=[None], mode='markers',
        marker=dict(
            colorscale='RdBu_r',
            cmin=min(sourceTargetDf['count']),
            cmax=max(sourceTargetDf['count']),
            colorbar=dict(
                title="Value",
                thickness=15,
                len=0.5,
                x=1.05,
                xref="paper"
            )
        ),
        hoverinfo='none'
    )

    fig.add_trace(colorbar_trace)

    fig.add_annotation(x=0, y=1.05, yref="paper", text="Source", showarrow=False, font=dict(size=10))
    fig.add_annotation(x=0.33, y=1.05, yref="paper", text="Ligand", showarrow=False, font=dict(size=10))
    fig.add_annotation(x=0.66, y=1.05, yref="paper", text="Receptor", showarrow=False, font=dict(size=10))
    fig.add_annotation(x=1, y=1.05, yref="paper", text="Target", showarrow=False, font=dict(size=10))

    fig.update_layout(
        xaxis=dict(showgrid=False, zeroline=False, showticklabels=False, range=[0,1]),
        yaxis=dict(showgrid=False, zeroline=False, showticklabels=False, range=[0,1]),
        plot_bgcolor='white',
        autosize=True,
        width = None,
        height = 600,
        title = title,
        font = dict(size=10)
        )
    
    fig.show(config={"responsive": True})

def gen_sankey(df, cat_cols=[], value_cols='', title='Sankey Diagram'):
    """
    Helper function to the function plot_sankey()

     Parameters
    ----------
    df :
        Dataframe

    cat_cols :
        Columns interested in the sankey plot

    value_cols :
        Sankey plot generated using connections based on this value_cols

    title :
        Title of Sankey plot
    
    Returns
    -------
    Nothing (plots Sankey plot)

    """

    # df['source'] += 'S'
    df['target'] += ' '
    
    labelList = []
    for catCol in cat_cols:
        labelListTemp =  list((df[catCol].values))
        labelList = labelList + labelListTemp    
        
    for i in range(len(cat_cols)-1):
        if i==0:
            sourceTargetDf = df[[cat_cols[i],cat_cols[i+1],value_cols]]
            sourceTargetDf.columns = ['source','target','count']
        else:
            tempDf = df[[cat_cols[i],cat_cols[i+1],value_cols]]
            tempDf.columns = ['source','target','count']
            sourceTargetDf = pd.concat([sourceTargetDf,tempDf])
    
    vmin = sourceTargetDf['count'].min()
    vmax = sourceTargetDf['count'].max()
    limit = max(abs(vmin), abs(vmax))
    vcenter = 0
    norm = mcolors.TwoSlopeNorm(vmin=-limit, vcenter=vcenter, vmax=limit)
    # norm = mcolors.Normalize(vmin=vmin, vmax=vmax)
    cmap = plt.get_cmap('RdBu_r')
    sourceTargetDf['hex_color'] = sourceTargetDf['count'].apply(lambda x: mcolors.to_hex(cmap(norm(x))))
    
    flows = []
    for i, row in sourceTargetDf.iterrows():
        flows.append((row['source'], row['target'], 1, {'color': row['hex_color']}))

    nodes = Sankey.infer_nodes(flows)
    nodes_new = []
    for level in nodes:
        level_new = []
        for node in level:
            node_new = node + [{'color' : 'black',
                                'label_pos':'center', 'label_opts': dict(fontsize=10, bbox=dict(boxstyle='round,pad=0.3', edgecolor='black', facecolor='white'))}]
            level_new.append(node_new)
        nodes_new.append(level_new)

    fig, ax = plt.subplots(figsize=(15, 10))
    s = Sankey(flows=flows,
               nodes=nodes_new,
               flow_color_mode_alpha=0.3,
               node_opts=dict(label_format='{label}'),
    )
    s.draw(ax=ax)

    sm = plt.cm.ScalarMappable(cmap=cmap, norm=norm)
    sm.set_array([])
    cbar = plt.colorbar(sm, ax=ax, orientation='vertical', pad=0.01, shrink=0.5)
    cbar.set_label(value_cols, fontsize=12)

    ax.text(x=-0.05, y=1.02, s="Source", fontsize=10)
    ax.text(x=0.95, y=1.02, s="Ligand", fontsize=10)
    ax.text(x=1.95, y=1.02, s="Receptor", fontsize=10)
    ax.text(x=2.95, y=1.02, s="Target", fontsize=10)

    ax.set_title(title)
    plt.tight_layout()
    plt.show()

def gene_annotation(gene_list_to_profile,
                    num_gos: int = 15,
                    figsize=(10,6),
                    title: str = None,
                    font_size: int = 10,
                    organism: str = 'hsapiens',
                    dpi: int = 100,
                    s: int = 100,
                    color: str = 'tab:blue'):
    """
    Perform Gene Ontology (GO) enrichment analysis and create a scatterplot of enriched terms.

    Parameters:
    ----------
    num_gos: int, optional
        Number of GO terms to plot. Default is 5.
    figsize: tuple, optional
        figsize. Default is (6,6).
    title: str
        Title of the plot.
    font_size: int, optional
        Font size for labels. Default is 10.
    0rganism: str, optional
        The organism for GO analysis. Default is 'hsapiens'.
    dpi: int, optional
        Dots per inch for the saved plot image. Default is 100.
    s: int, optional
        Marker size for the scatterplot. Default is 200.
    color: str, optional
        Color of the scatterplot markers. Default is 'tab:blue'.

    Returns:
    --------
    None
        Plots the scatterplot of enriched GO terms.
    """

    
    gp = GProfiler(return_dataframe=True)
    if gene_list_to_profile:
        gprofiler_results = gp.profile(organism = organism,
                                       query = gene_list_to_profile)
    else:
        return "Genes list is empty!"
    
    
    if(gprofiler_results.shape[0] == 0):
        return "Not enough information!"

    
    if(gprofiler_results.shape[0] < num_gos):
        num_gos = gprofiler_results.shape[0]

  
    selected_gps = gprofiler_results.head(num_gos)[['name', 'p_value']]
    
    selected_gps['nlog10'] = -np.log10(selected_gps['p_value'].values)

    plt.figure(figsize = figsize, dpi = dpi)
    # plt.style.use('default')
    sns.scatterplot(data = selected_gps, x = "nlog10", y = "name", s = s, color = color)

    plt.title(title, fontsize = font_size)

    plt.xticks(size = font_size)
    plt.yticks(size = font_size)

    plt.ylabel("GO Terms", size = font_size)
    plt.xlabel("-$log_{10}$ (P-value)", size = font_size)

    plt.tight_layout()
    plt.show()


def plot_volcane(df, method, p_threshold=0.05, fc_threshold=1, figsize=(8, 6), annot=True, title=None):
    """
    This function generates a Volcano plot

    Parameters
    ----------
    df :
        Dataframe

    Returns
    -------
    Python default volcano plot

    """
    np.random.seed(42)
    data = df
    data['neg_log10_p_value'] = -np.log10(df['p_value'])
    if method == "fisher":
        attr = "lodds"
    elif method == "mannwhitneyu":
        attr = "lfc"

    data["color"] = "gray"
    data.loc[(data[attr] > fc_threshold) & (data["p_value"] < p_threshold), "color"] = "red"
    data.loc[(data[attr] < -fc_threshold) & (data["p_value"] < p_threshold), "color"] = "red"
    data.loc[(data[attr] > -fc_threshold) & (data[attr] < fc_threshold) & (data["p_value"] < p_threshold), "color"] = "blue"
    data.loc[(data[attr] < -fc_threshold) & (data["p_value"] > p_threshold), "color"] = "green"
    data.loc[(data[attr] > fc_threshold) & (data["p_value"] > p_threshold), "color"] = "green"

    # Plot
    plt.figure(figsize=figsize)
    sns.scatterplot(x=attr, y="neg_log10_p_value", hue="color", palette={"gray": "gray", "red": "red", "blue": "blue", "green": "green"}, data=data, edgecolor=None, alpha=0.7)

    # Add significance threshold lines
    plt.axhline(-np.log10(p_threshold), linestyle="--", color="black", linewidth=1)
    plt.axvline(fc_threshold, linestyle="--", color="black", linewidth=1)
    plt.axvline(-fc_threshold, linestyle="--", color="black", linewidth=1)

    if annot:
        for i, row in data.iterrows():
            if row['color'] == 'red':
                plt.text(row[attr], row["neg_log10_p_value"], row["cellpair"], fontsize=8, ha='right')

    x_limit = max(abs(data[attr].min()), abs(data[attr].max()))
    plt.xlim(-x_limit-1, x_limit+1)

    plt.xlabel(r"Log$_{2}$ Fold Change")
    plt.ylabel(r"-Log$_{10}$(p-value)")
    plt.title(title)
    plt.legend([],[], frameon=False)
    plt.show()


def plot_clustermap(data, title, annot=True, return_figure=False):
    """
    This function generates a Clustermap plot

    Parameters
    ----------
    data :
        Dataframe
    title : str
        Title of the plot
    annot : bool
        Whether to annotate the heatmap with values
    return_figure:
        Option for return matplotlib figure axes
    Returns
    -------
    
    Python Cluster map

    """
    pivot_table = data.groupby(["source", "target"])["LRScore"].sum().unstack().fillna(0)
    xlabel, ylabel = "Target", "Source"
    g = sns.clustermap(
        pivot_table,
        figsize=(9, 7),
        annot=annot,
        linewidths=0.5,
        method="ward",
        metric="euclidean",
        dendrogram_ratio=(0.2, 0.2),
        cbar_pos=(0.02, 0.8, 0.03, 0.15)
    )
    g.ax_heatmap.set_xlabel(xlabel)
    g.ax_heatmap.set_ylabel(ylabel)
    plt.title(title, fontsize=14)
    plt.show()


def plot_graph_clustermap(graph, weight="LRScore", title="Ligand-Receptor Heatmap", annot=True):
    """
    This function generates the graph adjacency matrix Heatmap

    Parameters
    ----------
    data :
        Dataframe
    weight : str
        The weight attribute to use for the adjacency matrix, default is "LRScore"
    title : str
        Title of the plot
    annot : bool
        Whether to annotate the heatmap with values

    Returns
    -------
    Python Cluster map

    """
    graph = nx.from_pandas_edgelist(graph,
                                    source='source',
                                    target='target',
                                    edge_attr=True,
                                    create_using=nx.DiGraph())

    nodes = list(graph.nodes)
    adj_matrix = nx.to_pandas_adjacency(graph, nodelist=nodes, weight=weight).fillna(0).astype(float)
    max_val = np.abs(adj_matrix.values).max()

    g = sns.clustermap(
        adj_matrix,
        figsize=(9, 7),
        cmap="RdBu_r",
        linewidths=0.5,
        center=0,
        vmin=-max_val,
        vmax=max_val,
        annot=annot,
        method="ward",
        metric="euclidean",
        dendrogram_ratio=(0.2, 0.2),
        cbar_pos=(0.02, 0.8, 0.03, 0.15)
    )

    g.ax_heatmap.set_xlabel("Receptor Cluster")
    g.ax_heatmap.set_ylabel("Ligand Cluster")
    plt.title(title, fontsize=14)

    plt.show()