FBA solving got stuck with JSON format model

[lun-ai]

I noticed FBA got stuck occasionally and consistently on many gene deletions. I wonder what might lead to this.
I was using the yeast9 model (https://github.com/SysBioChalmers/yeast-GEM/tree/main).

FBA did not get stuck when I ran the gene deletion separately, and it finished in 0.6 seconds.
I saw a relevant question (#736), but this appears to be a different problem.

Edit:
I also encountered this when I ran the same code with the model in the SBML format. The JSON format model was returned from the save_json_model function. I left the code running overnight. It ran over 10 hours and could not finish, it took too long considering the individual FBA spent <0.01 secs.

Thank you.

================================================================================

I was running the below code:

from cobra.io import load_json_model, read_sbml_model, save_json_model

if __name__ == '__main__':

    model_path = 'AutoGEM/GEM/yeast9/yeast-GEM.json'
    model = load_json_model(model_path)
    gene_set = [gene.id for gene in model.genes]

    total_items = 0
    for i in range(len(gene_set)):
        g1 = gene_set[i]
        for j in range(i, len(gene_set)):
            g2 = gene_set[j]
            print(f"Gene {i}: {g1} - Gene {j}: {g2}")
            # Perform double gene deletion
            with model as inner:
                inner.genes.get_by_id(g1).knock_out()
                inner.genes.get_by_id(g2).knock_out()
                growth = inner.slim_optimize()
                total_items += 1
            print(f"Item No. {total_items}, result: {growth}")

The above hung at:

...
Time taken: 0.043855905532836914 seconds
Item No. 1553, result: 0.031474191119974834
Gene 1: Q0080 - Gene 393: YGL037C
Time taken: 0.008934497833251953 seconds
Item No. 1554, result: 0.03147419111997623
Gene 1: Q0080 - Gene 394: YGL038C
Time taken: 0.009885549545288086 seconds
Item No. 1555, result: 0.03147419111997623
Gene 1: Q0080 - Gene 395: YGL040C    <- took too long

I observed across different runs from the JSON format model that FBA got stuck at the same point:

Test1:

Test2:

System:

Architecture:             x86_64
  CPU op-mode(s):         32-bit, 64-bit
  Address sizes:          48 bits physical, 57 bits virtual
  Byte Order:             Little Endian
CPU(s):                   48
  On-line CPU(s) list:    0-47
Vendor ID:                GenuineIntel
  Model name:             Intel(R) Xeon(R) w7-3465X
    CPU family:           6
    Model:                143
    Thread(s) per core:   2
    Core(s) per socket:   24
    Socket(s):            1
    Stepping:             8
    BogoMIPS:             4991.99
    Flags:                fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ss ht syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon r
                          ep_good nopl xtopology tsc_reliable nonstop_tsc cpuid pni pclmulqdq vmx ssse3 fma cx16 pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave 
                          avx f16c rdrand hypervisor lahf_lm abm 3dnowprefetch invpcid_single ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow vnmi ept vpid ept_ad fsgsbase tsc_adjust bmi1 
                          avx2 smep bmi2 erms invpcid avx512f avx512dq rdseed adx smap avx512ifma clflushopt clwb avx512cd sha_ni avx512bw avx512vl xsaveopt xsavec xgetbv1 xsaves avx_vnn
                          i avx512_bf16 avx512vbmi umip waitpkg avx512_vbmi2 gfni vaes vpclmulqdq avx512_vnni avx512_bitalg avx512_vpopcntdq rdpid cldemote movdiri movdir64b fsrm md_clea
                          r serialize tsxldtrk arch_lbr avx512_fp16 flush_l1d arch_capabilities
Virtualization features:  
  Virtualization:         VT-x
  Hypervisor vendor:      Microsoft
  Virtualization type:    full
Caches (sum of all):      
  L1d:                    1.1 MiB (24 instances)
  L1i:                    768 KiB (24 instances)
  L2:                     48 MiB (24 instances)
  L3:                     75 MiB (1 instance)
Vulnerabilities:          
  Gather data sampling:   Not affected
  Itlb multihit:          Not affected
  L1tf:                   Not affected
  Mds:                    Not affected
  Meltdown:               Not affected
  Mmio stale data:        Not affected
  Reg file data sampling: Not affected
  Retbleed:               Mitigation; Enhanced IBRS
  Spec rstack overflow:   Not affected
  Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl and seccomp
  Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization
  Spectre v2:             Mitigation; Enhanced / Automatic IBRS; IBPB conditional; RSB filling; PBRSB-eIBRS SW sequence; BHI BHI_DIS_S
  Srbds:                  Not affected
  Tsx async abort:        Not affected

[lun-ai]

I tried with the Gurobi solver, and the problem was resolved. I think it might just be an issue of the default solver.

[cdiener]

Yeah I agree it was probably GLPK getting stuck in an invalid basis. Sometimes turning presolving on or off can help with that.

[lun-ai]

This makes a lot of sense. Could you suggest how to switch this "presolving" flag? I had a quick look online and did not find a page about it.

[cdiener]

Sure, you set all solver parameters via Model.solver.configuration. For instance my_model.solver.configuration.presolve = True.