HiGHS: MIP yields incorrect result with presolve on.

This is one of just a few solutions that are incorrect. Perhaps there is an option I can change?

subset_select_highs_in.mps.txt

Exact solution: Optimal = 44

With presolve on:

Solving report
  Status            Optimal
  Primal bound      45
  Dual bound        45
  Gap               0% (tolerance: 0.01%)
  Solution status   feasible
                    45 (objective)
                    0 (bound viol.)
                    0 (int. viol.)
                    0 (row viol.)
  Timing            0.01 (total)
                    0.00 (presolve)
                    0.00 (postsolve)
  Nodes             0
  LP iterations     21 (total)
                    0 (strong br.)
                    3 (separation)
                    0 (heuristics)

With presolve off:

Solving report
  Status            Optimal
  Primal bound      44
  Dual bound        44
  Gap               0% (tolerance: 0.01%)
  Solution status   feasible
                    44 (objective)
                    0 (bound viol.)
                    2.23376872555e-13 (int. viol.)
                    0 (row viol.)
  Timing            0.01 (total)
                    0.00 (presolve)
                    0.00 (postsolve)
  Nodes             1
  LP iterations     51 (total)
                    0 (strong br.)
                    23 (separation)
                    0 (heuristics)

subset_select_job3_highs-presolve-off.log subset_select_job3_highs-presolve-on.log

I only saw one LP option and these MIP options. Would one of them be appropriate?

  // Options for MIP solver
  bool mip_detect_symmetry;
  HighsInt mip_max_nodes;
  HighsInt mip_max_stall_nodes;
  HighsInt mip_max_leaves;
  HighsInt mip_max_improving_sols;
  HighsInt mip_lp_age_limit;
  HighsInt mip_pool_age_limit;
  HighsInt mip_pool_soft_limit;
  HighsInt mip_pscost_minreliable;
  HighsInt mip_min_cliquetable_entries_for_parallelism;
  HighsInt mip_report_level;
  double mip_feasibility_tolerance;
  double mip_rel_gap;
  double mip_abs_gap;
  double mip_heuristic_effort;