NetworkX: drawing large graphs?

I have a graphML file that represents relation between multiple SQL tables (nodes and edges).

I am using networkx to parse the file and matplotlib to draw it. The issue I have is that my graph is quite big (around 150 nodes) and hard to read. I do not have any experience in computer graph / drawing and the final result looks messy, is there a way to improve the readability of the graph (another library for instance, larger image,...) without decreasing the total number of nodes?

import networkx as nx
import matplotlib.pyplot as plt

        
input_graph = nx.read_graphml("graph.graphml")
    
to_remove = []
for node, data in input_graph.nodes(data=True):
    if data['zone'] != 'gold' or input_graph.degree(node) == 0: 
       to_remove.append(node)
    
input_graph.remove_nodes_from(to_remove)
nx.draw(input_graph, with_labels=True)
plt.show()

Some of the options are:

• shorten node labels and make use of the colours to differentiate groups/categories of nodes (e.g. nodes that belong to group A are coloured with red, nodes that belong to B are blue, etc.);

• experiment with the layout: the default layout is spring layout, but another layout might give you less clutter, e.g. circular layout;

• use datashader's edge bundling.